INDEX

Explanations

instances of the word "look" in various contexts

oai_token-act-pair · gpt-3.5-turbo

phrases that discuss visual appearances or perspectives

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

 Literature

-0.70

Staff

-0.68

actionDate

-0.66

EStreamFrame

-0.66

 Kurd

-0.64

ommod

-0.63

cffff

-0.63

 Loch

-0.63

CENT

-0.62

KEN

-0.62

POSITIVE LOGITS

ahead

1.07

checks

0.86

finder

0.82

outs

0.81

lessly

0.78

ENE

0.76

ups

0.76

behind

0.74

anka

0.73

apeshifter

0.72

Activations Density 0.025%

instances of the word "look" in various contexts

phrases that discuss visual appearances or perspectives

No Comments

No Known Activations