INDEX

Explanations

emphatic expressions used to convey strong opinions or judgments

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

outright or downright negative states

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

colgante

-0.54

writeFieldEnd

-0.51

 дописавши

-0.50

MockBean

-0.49

 unknownFields

-0.48

niña

-0.47

 ainfi

-0.47

 présidenti

-0.46

niño

-0.45

 useAppContext

-0.45

POSITIVE LOGITS

 outright

0.96

 downright

0.93

 blatantly

0.71

 blatant

0.70

 pure

0.65

 whopping

0.64

 sheer

0.60

 utterly

0.59

pure

0.59

 freaking

0.58

Activations Density 0.018%

emphatic expressions used to convey strong opinions or judgments

outright or downright negative states

No Comments

No Known Activations