INDEX

Explanations

questions and statements regarding legality and agreements

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

Tokens preceding listed words

np_acts-logits-general · gemini-2.0-flash

thousand cuts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

 ſch

-0.78

 iſt

-0.74

 cauſe

-0.73

 Guarda

-0.72

 appelle

-0.71

 purpoſe

-0.71

 myſelf

-0.71

 ſtate

-0.70

 Trigonometry

-0.70

 faſt

-0.69

POSITIVE LOGITS

Re

0.73

Dis

0.67

En

0.64

Sub

0.64

Res

0.62

Pre

0.61

Pro

0.61

Sp

0.60

Ste

0.60

Un

0.60

Activations Density 0.856%

questions and statements regarding legality and agreements

Tokens preceding listed words

thousand cuts

No Comments

No Known Activations