INDEX

Explanations

references to physical alterations or severances

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

cutting or severing

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

stylers

-0.47

 undertook

-0.44

LEncoder

-0.44

Дереккөздер

-0.43

generally

-0.42

 ouder

-0.40

 understandings

-0.39

wikkeld

-0.39

USF

-0.39

Generally

-0.38

POSITIVE LOGITS

cut

0.73

 Cutting

0.66

 cutting

0.66

CUT

0.64

Cut

0.62

 potong

0.62

Cutting

0.62

cut

0.61

 cuts

0.59

Cut

0.58

Activations Density 0.039%

references to physical alterations or severances

cutting or severing

No Comments

No Known Activations