INDEX

Explanations

content related to news articles or reports

oai_token-act-pair · gpt-3.5-turbo

segments where self-reflection or personal experiences are discussed

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 blinded

-0.75

 undermin

-0.73

 engagements

-0.71

 footing

-0.70

 wiser

-0.69

 proport

-0.68

 capit

-0.68

everal

-0.65

 composing

-0.65

 endeavour

-0.65

POSITIVE LOGITS

AMY

0.97

MJ

0.92

âĹ¼

0.88

ÂłÂłÂłÂłÂłÂłÂłÂł

0.83

POST

0.82

laughs

0.80

NER

0.80

Interview

0.80

Anyway

0.78

JC

0.78

Activations Density 0.281%

content related to news articles or reports

segments where self-reflection or personal experiences are discussed

No Known Activations