INDEX

Explanations

instances of hostility or conflict

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Configuration

jbloom/Gemma-2b-IT-Residual-Stream-SAEs/gemma_2b_it_blocks.12.hook_resid_post_16384

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

HuggingFaceFW/fineweb

Features

16,384

Data Type

float32

Hook Name

blocks.12.hook_resid_post

Hook Layer

Architecture

standard

Context Size

1,024

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 carrefour

-0.61

 métier

-0.58

 moulin

-0.57

 célé

-0.56

 loup

-0.54

 PLATES

-0.54

 cahier

-0.54

 Româ

-0.53

 vernac

-0.53

 Chinois

-0.53

POSITIVE LOGITS

 veniva

0.86

Solución

0.84

 faceva

0.82

 intende

0.78

IsContent

0.76

 patrio

0.75

 nuovamente

0.74

 kriminal

0.73

 perciò

0.73

ⓧ

0.72

Activations Density 0.406%

instances of hostility or conflict

No Comments

No Known Activations