INDEX

Explanations

positive adjectives that indicate value or benefits

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

helpful, useful, and positive attributes

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

 snippetHide

-0.59

rungsseite

-0.57

ceral

-0.55

enterOuterAlt

-0.54

verwijspagina

-0.52

IntoConstraints

-0.52

setVerticalGroup

-0.51

 '\\;'

-0.51

fromnode

-0.51

dafx

-0.50

POSITIVE LOGITS

 helpful

0.57

 useful

0.54

 invaluable

0.52

 handy

0.52

 usefulness

0.51

 convenience

0.50

 tool

0.50

 invention

0.50

 discovery

0.48

 tools

0.47

Activations Density 0.037%

positive adjectives that indicate value or benefits

helpful, useful, and positive attributes

No Comments

No Known Activations