INDEX

Explanations

instances of parentheses or function calls

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

listing items

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

transQ

-0.54

 ligiloj

-0.52

RegressionTest

-0.49

 propOrder

-0.46

 GenerationType

-0.44

RTDA

-0.44

 mxArray

-0.44

InitVars

-0.43

bufio

-0.42

setof

-0.42

POSITIVE LOGITS

 unique

0.50

 descar

0.47

时候

0.46

 Chow

0.45

 representative

0.45

++]=

0.45

return

0.45

 correct

0.44

the

0.44

ioren

0.44

Activations Density 0.010%

instances of parentheses or function calls

listing items

No Comments

No Known Activations