INDEX

Explanations

phrases indicating opinions or evaluations

oai_token-act-pair · gpt-3.5-turbo

negations or phrases expressing the opposite of a statement

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

"(

-0.56

thinkable

-0.54

IRE

-0.54

Is

-0.54

Assembly

-0.53

osi

-0.52

ahs

-0.51

 bathing

-0.51

'd

-0.50

idian

-0.50

POSITIVE LOGITS

apego

0.67

 itself

0.66

 acronym

0.65

retty

0.61

 decentralized

0.60

 decentral

0.58

 disbanded

0.58

 phased

0.57

igmatic

0.57

 Balk

0.55

Activations Density 1.062%

phrases indicating opinions or evaluations

negations or phrases expressing the opposite of a statement

No Known Activations