INDEX

Explanations

negative assessments related to laws or systems

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

French "mauvaise" (bad) and related words

np_acts-logits-general · gemini-2.0-flash

responses and answers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_16k/average_l0_88

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.9.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

 nahilalakip

-0.54

RTEX

-0.53

Tikang

-0.51

 ligiloj

-0.46

WriteTagHelper

-0.45

 transfieras

-0.42

 '\\;'

-0.42

 propOrder

-0.42

addGap

-0.41

 fieldNum

-0.40

POSITIVE LOGITS

worst

0.72

 worst

0.70

 mauvais

0.66

 horrible

0.65

Worst

0.63

 buruk

0.59

bad

0.59

 mauvaises

0.58

 Worst

0.58

 mauvaise

0.58

Activations Density 1.748%

negative assessments related to laws or systems

French "mauvaise" (bad) and related words

responses and answers

No Comments

No Known Activations