INDEX

Explanations

terms related to political or government actions and responses

oai_token-act-pair · gpt-3.5-turbo

terms related to governance and accountability in political contexts

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 myself

-0.71

 oneself

-0.67

oke

-0.65

 Corpor

-0.65

 Brunswick

-0.63

ÃŁ

-0.61

 Anon

-0.59

glomer

-0.58

kan

-0.56

ga

-0.56

POSITIVE LOGITS

 abroad

0.84

reach

0.79

iness

0.78

 foray

0.75

 regarding

0.73

 toward

0.73

 towards

0.73

vention

0.73

 woes

0.72

worthiness

0.72

Activations Density 0.348%

terms related to political or government actions and responses

terms related to governance and accountability in political contexts

No Comments

No Known Activations