INDEX

Explanations

phrases where someone is admitting something

oai_token-act-pair · gpt-3.5-turbo

instances of the word "admit."

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

ammy

-0.81

ccording

-0.73

chn

-0.71

srf

-0.69

miah

-0.68

rior

-0.67

 impacted

-0.66

eled

-0.66

phabet

-0.64

gm

-0.63

POSITIVE LOGITS

ibility

0.97

 admit

0.93

IBLE

0.84

 admitting

0.83

 admits

0.77

 admission

0.75

 admitted

0.75

iary

0.75

essors

0.73

ably

0.72

Activations Density 0.014%

phrases where someone is admitting something

instances of the word "admit."

No Comments

No Known Activations