INDEX

Explanations

phrases indicating something that has not been done or completed yet

oai_token-act-pair · gpt-3.5-turbo

phrases indicating something that has not occurred or been completed yet

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

ufact

-0.82

gang

-0.69

 similarities

-0.65

ãĥ¼ãĥ

-0.65

chio

-0.62

ãĥ³ãĤ¸

-0.60

 disproportion

-0.60

ging

-0.60

packs

-0.59

gers

-0.59

POSITIVE LOGITS

terday

0.78

?:

0.72

hin

0.71

;)

0.70

 anyways

0.70

here

0.69

 anyway

0.69

NESS

0.69

!!

0.68

Activations Density 0.028%

phrases indicating something that has not been done or completed yet

phrases indicating something that has not occurred or been completed yet

No Comments

No Known Activations