INDEX

Explanations

code structure and programming constructs, particularly function and variable definitions, assertions, and memory management

oai_token-act-pair · gpt-4o-mini Triggered by @bot

Relational and equality operators

np_acts-logits-general · gemini-2.0-flash

NULL comparisons to invariant

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 CGFloat

-0.64

[];

-0.59

 null

-0.58

'':

-0.58

cor

-0.57

''

-0.56

 Math

-0.56

=["

-0.56

 boolean

-0.55

[]

-0.54

POSITIVE LOGITS

 vetorial

0.85



0.78

static

0.75

 vectoriales

0.75

 feroit

0.75

 pleaſure

0.73

 auroit

0.72

 literaria

0.71

 literario

0.70

 egip

0.69

Activations Density 0.349%

code structure and programming constructs, particularly function and variable definitions, assertions, and memory management

Relational and equality operators

NULL comparisons to invariant

No Comments

No Known Activations