INDEX

Explanations

mentions of media-related statements

oai_token-act-pair · gpt-3.5-turbo

references to media and press releases

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

thal

-0.80

orney

-0.74

owicz

-0.69

kun

-0.66

 polyg

-0.66

innon

-0.64

atars

-0.64

 Gustav

-0.63

esar

-0.63

dies

-0.62

POSITIVE LOGITS

 playback

1.02

 Availability

0.83

 Release

0.83

 Editors

0.82

 RELEASE

0.77

 Coverage

0.76

 Releases

0.76

eval

0.76

Reference

0.76

 RELE

0.75

Activations Density 0.030%

mentions of media-related statements

references to media and press releases

No Comments

No Known Activations