INDEX

Explanations

references to geographical features, specifically cliffs

oai_token-act-pair · gpt-3.5-turbo

mentions of cliffs

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Role

-0.90

Interstitial

-0.82

rity

-0.82

alg

-0.74

apolis

-0.71

ADRA

-0.71

enza

-0.70

ulton

-0.70

aniel

-0.68

National

-0.68

POSITIVE LOGITS

 cliff

1.15

 cliffs

1.09

 ledge

0.88

 slope

0.83

 bluff

0.83

 dwellings

0.82

frog

0.79

 climbs

0.79

 waterfall

0.79

 concessions

0.78

Activations Density 0.007%

references to geographical features, specifically cliffs

mentions of cliffs

No Comments

No Known Activations