INDEX

Explanations

written documents such as memos or letters

oai_token-act-pair · gpt-3.5-turbo

mentions of memoranda or official documents

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

reek

-0.73

Fen

-0.70

rates

-0.65

places

-0.63

lihood

-0.63

Pyr

-0.61

 Greeks

-0.61

tool

-0.60

alien

-0.59

respect

-0.59

POSITIVE LOGITS

andum

1.08

 memo

1.07

 memos

0.95

ographed

0.85

ufact

0.84

isode

0.83

ariat

0.83

velop

0.82

ovie

0.81

osal

0.80

Activations Density 0.014%

written documents such as memos or letters

mentions of memoranda or official documents

No Comments

No Known Activations