INDEX

Explanations

nouns related to large accumulations or groupings of objects

oai_token-act-pair · gpt-3.5-turbo

references to "piles" in various contexts

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

saf

-0.69

oral

-0.66

hips

-0.66

Interstitial

-0.63

otiation

-0.62

javascript

-0.62

 obser

-0.61

 livest

-0.61

avid

-0.60

POSITIVE LOGITS

 piles

1.09

 pile

1.07

 piling

1.04

bags

0.99

 piled

0.95

bag

0.87

wagon

0.86

endium

0.80

 heap

0.76

aging

0.75

Activations Density 0.013%

nouns related to large accumulations or groupings of objects

references to "piles" in various contexts

No Comments

No Known Activations