INDEX

Explanations

code-related elements particularly focused on header files and libraries

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

programming language constructs

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 queſta

-1.05

writeFieldEnd

-0.98

ConstraintMaker

-0.96

IntoConstraints

-0.94

enderror

-0.94

adaptiveStyles

-0.94

+#+

-0.94

ngrx

-0.93

twimg

-0.91

ftagPool

-0.90

POSITIVE LOGITS

0.46

import

0.45

0.42

0.40

But

0.40

En

0.39

The

0.38

Un

0.36

0.35

 но

0.35

Activations Density 0.014%

code-related elements particularly focused on header files and libraries

programming language constructs

No Comments

No Known Activations