INDEX

Explanations

references to specific locations, like "Capitol Hill"

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

fuf

-0.92

 michelin

-0.90

 encomp

-0.89

 scrat

-0.88

 increa

-0.87

 nutella

-0.85

 affor

-0.85

 guarante

-0.85

 tiffany

-0.85

 oreo

-0.83

POSITIVE LOGITS

 Capitol

1.42

Capitol

1.16

 capitol

0.94

 Capital

0.79

 lawmakers

0.66

Capital

0.65

 Congressional

0.62

 Kapit

0.60

 capital

0.60

 Congress

0.59

Activations Density 0.061%

references to specific locations, like "Capitol Hill"

No Comments

No Known Activations