INDEX

Explanations

names of people involved in editing or reporting in a professional context

oai_token-act-pair · gpt-3.5-turbo

sections indicating authorship or editing credits in a document

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

gdala

-0.86

ochond

-0.84

worthiness

-0.81

egu

-0.80

omorphic

-0.77

bucks

-0.76

objects

-0.75

ulnerable

-0.73

omet

-0.73

incinn

-0.72

POSITIVE LOGITS

TBA

0.89

 Gareth

0.89

 Johannes

0.85

 Brent

0.83

 Tomas

0.82

 Rafael

0.82

 Deborah

0.81

 Robert

0.80

 Geoff

0.80

 William

0.79

Activations Density 0.190%

names of people involved in editing or reporting in a professional context

sections indicating authorship or editing credits in a document

No Comments

No Known Activations