INDEX

Explanations

references related to the name "Obama"

oai_token-act-pair · gpt-3.5-turbo

mentions of Obama and related political contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

Cth

-0.84

 ILCS

-0.82

itionally

-0.81

asma

-0.71

ounge

-0.71

asus

-0.69

amel

-0.69

 Edinburgh

-0.69

 Newcastle

-0.68

TOR

-0.68

POSITIVE LOGITS

 administration

1.13

Care

1.07

 Administration

1.07

Obama

0.95

omics

0.89

 appoint

0.88

 Admin

0.88

 advisers

0.87

 reelection

0.87

 adviser

0.86

Activations Density 0.050%

references related to the name "Obama"

mentions of Obama and related political contexts

No Comments

No Known Activations