Exponent
An exponent is a number that specifies the power to which a base is raised, including the radix power that scales a significand in floating-point representation.
JavaScript in Depth
A value specifying the power to which a radix is raised to scale a significand.
- How it is used
- Explains how an exponent scales a significand in floating-point number encoding.
- How deep it goes
- Defines the exponent and works through its 11-bit biased Binary64 representation.
- Examples
- Uses 12345 × 10⁻² and the binary encodings of 0.1 and 0.2.
- What you can do after
- Evaluate/understand how stored exponents determine floating-point scale and why addition requires exponent alignment.
Rust in Action: Systems programming concepts and techniques
An exponent is the component of a floating-point representation that determines the scale or power of two applied to its significand.
- How it is used
- Uses the exponent to explain floating-point scale and control the range of values constructed from bytes.
- How deep it goes
- Works through biased exponent extraction and decoding, then identifies special exponent patterns that change normal interpretation.
- Examples
- Decodes the exponent of 42.42 and constructs floating-point values from a byte using a fixed exponent.
- What you can do after
- Apply it via Rust bit operations to extract an f32 exponent or construct a bounded floating-point value.
Read how Rust in Action: Systems programming concepts and techniques explains exponent →
Quantization and Fast Inference
The scale-bearing component of a floating-point representation that determines the magnitude of its significand.
- How it is used
- Explains how floating-point scales significands across magnitudes and contrasts adaptive range with fixed-point grids.
- How deep it goes
- Defines value as mantissa × 2^exponent, illustrates a toy format numerically, and stops short of IEEE-754 encoding.
- Examples
- Uses 0.01, 0.10, 1.00, and 3.00 in a toy float to show exponent-dependent spacing.
- What you can do after
- Understand how an exponent moves the representational grid to provide scale-dependent resolution.
Read how Quantization and Fast Inference explains exponent →
100 Go Mistakes and How to Avoid Them
An exponent is the multiplier component that determines the power applied to a floating-point mantissa.
- How it is used
- Explains floating-point storage and the precision loss developers observe in calculations.
- How deep it goes
- Defines exponent bits and their role in the representation, but stops short of explaining conversion mechanics.
- Examples
- Uses the float32 encoding of 1.0001 and the formula sign × 2^exponent × mantissa.
- What you can do after
- Evaluate/understand why a floating-point value can differ from its written decimal form.
Read how 100 Go Mistakes and How to Avoid Them explains exponent →
Build a DeepSeek Model (From Scratch)
An exponent is the component of a floating-point representation that determines the scale and dynamic range of represented values.
- How it is used
- Explains how floating-point formats control numerical range and why quantization designs allocate bits between exponents and mantissas.
- How deep it goes
- Defines exponent bits, compares FP32, FP16, BF16, E5M2, and E4M3, then applies the trade-off to fine-grained quantization.
- Examples
- Compares FP32, FP16, BF16, E5M2, and E4M3 through their exponent-mantissa allocations and training roles.
- What you can do after
- Evaluate numerical formats by predicting their dynamic range, precision, and susceptibility to overflow or underflow.
Read how Build a DeepSeek Model (From Scratch) explains exponent →
CUDA for LLMs
A floating-point exponent is the field that determines the order of magnitude of a represented number.
- How it is used
- Uses the exponent–significand split to explain how floating-point formats represent magnitude and precision in quantization.
- How deep it goes
- Defines the exponent’s role and contrasts fixed format fields with adjustable integer ranges, without detailing bit layouts.
- Examples
- Compares FP8, FP16, and FP4 formats with integer quantization and external scaling.
- What you can do after
- Evaluate why low-precision floating-point formats need external scales to adjust range for data distributions.
Also covered in
- Modern C, Third Edition
- Modern C
- Deep Learning with R, Third Edition
- Grokking Machine Learning, Second EditionIn early access (MEAP)
- Math for Web Design
What to know first
- bitA bit is a binary digit and unit of digital information that represents one of two distinguishable values, conventionally 0 or 1.
- float64Float64 is a 64-bit floating-point numeric type, commonly represented by the IEEE 754 binary64 format.
- floating-point arithmeticFloating-point arithmetic is numerical computation using finite-precision representations of real numbers, which can introduce rounding and range errors.
- IEEE 754IEEE 754 is an international standard specifying floating-point formats, arithmetic operations, rounding rules, and exceptional values.
Related concepts
- floating-point arithmeticpart of · Floating-point arithmetic is numerical computation using finite-precision representations of real numbers, which can introduce rounding and range errors.
- power functionused by · Exponentiation is the mathematical operation of raising a base to a specified exponent or power.
- floating-point valuepart of · A floating-point value is a finite-precision approximation of a real number, typically represented using a significand and an exponent.
- damping exponentincludes · A damping exponent is a nonnegative exponent that controls how graph degree values reduce the contribution of a path.
Questions about exponent
What is exponent?
An exponent is a number that specifies the power to which a base is raised, including the radix power that scales a significand in floating-point representation.
Which Manning books explain exponent?
JavaScript in Depth, Rust in Action: Systems programming concepts and techniques, Quantization and Fast Inference, 100 Go Mistakes and How to Avoid Them, Build a DeepSeek Model (From Scratch) and CUDA for LLMs.
What should I know before learning exponent?
Manning books that teach exponent usually introduce bit, float64, floating-point arithmetic and IEEE 754 first.
What kinds of exponent are there?
Kinds of exponent include damping exponent and scaling exponent.
What is exponent part of?
Exponent is part of floating-point arithmetic. Floating-point arithmetic is numerical computation using finite-precision representations of real numbers, which can introduce rounding and range errors.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.





