INDEX

Explanations

requests for additional information or content

oai_token-act-pair · gpt-3.5-turbo

content indicating additional information or further reading

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

 Kodi

-0.63

 Patel

-0.61

Mun

-0.59

 Supervisor

-0.58

 Miliband

-0.58

 Himal

-0.57

 Leaf

-0.57

BST

-0.56

 Alpine

-0.56

 Mountain

-0.56

POSITIVE LOGITS

anwhile

0.92

interrupted

0.88

 dunno

0.84

etheless

0.78

[/

0.75

iott

0.75

wait

0.73

................

0.71

 christ

0.69

Wait

0.68

Activations Density 0.170%

requests for additional information or content

content indicating additional information or further reading

No Comments

No Known Activations