INDEX

Explanations

words related to physical collisions or impacts

oai_token-act-pair · gpt-3.5-turbo

references to collisions and related incidents

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

arag

-0.83

estate

-0.83

ifter

-0.81

ibaba

-0.79

zel

-0.79

iciency

-0.78

endment

-0.77

azy

-0.76

iatric

-0.76

zee

-0.75

POSITIVE LOGITS

 collisions

0.88

 collision

0.78

 shenan

0.76

 force

0.75

 Odyssey

0.74

 Franch

0.72

SHIP

0.72

worthiness

0.71

engers

0.71

halla

0.70

Activations Density 0.017%

words related to physical collisions or impacts

references to collisions and related incidents

No Comments

No Known Activations