INDEX

Explanations

the word "thing" with a high level of activation

oai_token-act-pair · gpt-3.5-turbo

references to a central concept or idea described as "thing."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

inav

-0.88

 anonymity

-0.66

cling

-0.66

ilings

-0.66

cled

-0.65

ardi

-0.65

oufl

-0.64

 largeDownload

-0.64

hurst

-0.63

cul

-0.63

POSITIVE LOGITS

iverse

0.93

Else

0.89

 happened

0.89

 happening

0.89

 happens

0.88

 happ

0.87

 Valiant

0.84

 thing

0.76

 happen

0.72

 else

0.72

Activations Density 0.034%

the word "thing" with a high level of activation

references to a central concept or idea described as "thing."

No Comments

No Known Activations

the word "thing" with a high level of activation

references to a central concept or idea described as "thing."

No Comments

No Known Activations