INDEX

Explanations

references to physical objects or entities

oai_token-act-pair · gpt-3.5-turbo

references to physical objects or items

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

gd

-0.78

quist

-0.75

glers

-0.75

bare

-0.74

adish

-0.74

ered

-0.71

atory

-0.70

emi

-0.69

shire

-0.69

linger

-0.69

POSITIVE LOGITS

ity

0.98

 manifestation

0.94

 sciences

0.91

ized

0.88

 impossibility

0.88

isation

0.85

anguage

0.83

IZE

0.83

ITY

0.81

ities

0.80

Activations Density 0.023%

references to physical objects or entities

references to physical objects or items

No Comments

No Known Activations