INDEX

Explanations

mentions of political leaders or officials, particularly referring to a "prime minister"

oai_token-act-pair · gpt-3.5-turbo

mentions of political leaders and their titles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

edIn

-0.79

 Locations

-0.79

 Warfare

-0.70

 Condition

-0.69

 Local

-0.68

ourcing

-0.66

 Contents

-0.66

 Latter

-0.65

 Blend

-0.64

 Casting

-0.64

POSITIVE LOGITS

's

0.97

 himself

0.96

ÃŃs

0.89

 vetoed

0.85

 congratulated

0.84

alty

0.82

 apologised

0.82

 apologized

0.78

 personally

0.77

 statue

0.77

Activations Density 0.145%

mentions of political leaders or officials, particularly referring to a "prime minister"

mentions of political leaders and their titles

No Comments

No Known Activations