INDEX

Explanations

references to walls

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

wall and walls

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

Tikang

-0.54

CBC

-0.49

mente

-0.46

ide

-0.46

Bec

-0.45

ised

-0.45

ly

-0.44

ActionPerformed

-0.43

Cee

-0.43



-0.42

POSITIVE LOGITS

 wall

2.13

Wall

1.98

wall

1.92

 Wall

1.89

 WALL

1.79

WALL

1.73

 walls

1.62

 Walls

1.53

walls

1.52

Walls

1.51

Activations Density 0.006%

references to walls

wall and walls

No Comments

No Known Activations

references to walls

wall and walls

No Comments

No Known Activations