INDEX

Explanations

references to Buffalo and its associated locations or events

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

buffalo and bison

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

subsection

-0.45

ter

-0.43

 περ

-0.41

rt

-0.40

denken

-0.40

hr

-0.40

ly

-0.39

-0.38

 περί

-0.38

lt

-0.38

POSITIVE LOGITS

 Buffalo

1.34

Buffalo

1.27

 buffalo

1.21

buffalo

1.14

 BUFF

0.90

 bison

0.77

 Buff

0.76

BUFF

0.72

 BUFFER

0.70

 themſelves

0.70

Activations Density 0.005%

references to Buffalo and its associated locations or events

buffalo and bison

No Comments

No Known Activations