INDEX

Explanations

connections and interactions in various contexts

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

comparing differences and statistics

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

 ModelExpression

-0.84

帖最后由

-0.69

 تضيفلها

-0.66

 <<<<<<<<<<<<<<

-0.66

 disambiguazione

-0.66

 NSCoder

-0.64

MigrationBuilder

-0.63

ագրություններ

-0.63

AndEndTag

-0.63

✨:

-0.61

POSITIVE LOGITS

↵↵

0.30

뀐

0.29

bumptech

0.29

 reduced

0.28

起

0.27

平

0.27

 scan

0.26

 mất

0.26

 ordinary

0.26

Activations Density 0.076%

connections and interactions in various contexts

comparing differences and statistics

No Comments

No Known Activations