INDEX

Explanations

specific titles or positions of political figures

oai_token-act-pair · gpt-3.5-turbo

terms related to political roles and titles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ologically

-0.73

sis

-0.73

iesta

-0.68

schild

-0.67

 psychologically

-0.64

zb

-0.64

condition

-0.63

phal

-0.63

anco

-0.63

zsche

-0.63

POSITIVE LOGITS

 Barth

0.80

utenant

0.77

 Franch

0.76

 Gerry

0.72

 Theodore

0.72

rict

0.72

 Died

0.71

 Ernst

0.71

aunt

0.69

 resolutions

0.68

Activations Density 0.053%

specific titles or positions of political figures

terms related to political roles and titles

No Comments

No Known Activations