INDEX

Explanations

questions that have been posed to individuals during interviews or public appearances

oai_token-act-pair · gpt-3.5-turbo

instances of the word "asked."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

marine

-0.73

 Scouting

-0.69

 Revision

-0.68

audi

-0.65

 Cock

-0.65

clude

-0.63

 Tigers

-0.61

redo

-0.61

Charge

-0.60

east

-0.60

POSITIVE LOGITS

 rhet

1.24

 asked

0.99

 questions

0.93

ioned

0.90

 questioned

0.85

 probing

0.85

 asks

0.82

 moderator

0.82

naires

0.81

govtrack

0.81

Activations Density 0.024%

questions that have been posed to individuals during interviews or public appearances

instances of the word "asked."

No Comments

No Known Activations