INDEX

Explanations

references to media or information sources

oai_token-act-pair · gpt-3.5-turbo

mentions of media

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

thal

-0.70

kun

-0.69

bats

-0.69

abouts

-0.68

atars

-0.67

itudinal

-0.66

ukong

-0.64

ãĤ¡

-0.63

chest

-0.62

affles

-0.62

POSITIVE LOGITS

 playback

1.03

eval

0.99

 outlets

0.86

 Release

0.80

 Coverage

0.77

 wishing

0.76

 RELEASE

0.74

eals

0.71

 Availability

0.71

 Matters

0.71

Activations Density 0.024%

references to media or information sources

mentions of media

No Comments

No Known Activations