INDEX

Explanations

specific phrases and structures that indicate questioning or clarification of significance and value

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

not what

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

HasAnnotation

-0.40

freiheit

-0.38

ItemBackground

-0.35

 муніципалі

-0.35

artiges

-0.35

qid

-0.34

 bestehen

-0.33

 fopen

-0.32

 Putih

-0.32

POSITIVE LOGITS

 Италијани

0.67

URLException

0.66

rungsseite

0.63

最快更新

0.60

ांकि

0.58

 SafeMath

0.58

necessarily

0.57

ThroughAttribute

0.56

culoskeletal

0.56

 necessarily

0.56

Activations Density 0.032%

specific phrases and structures that indicate questioning or clarification of significance and value

not what

No Comments

No Known Activations