INDEX

Explanations

phrases showing confusion or being confused

oai_token-act-pair · gpt-3.5-turbo

mentions of confusion and related states of uncertainty

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

ector

-0.69

bors

-0.64

tsky

-0.64

 Janeiro

-0.63

Reviewer

-0.63

baugh

-0.63

ACH

-0.61

OHN

-0.61

iaries

-0.61

uner

-0.60

POSITIVE LOGITS

 terminology

0.85

 ABOUT

0.80

ingly

0.79

 meanings

0.78

 confuse

0.77

 between

0.76

 semantics

0.76

otomy

0.74

 notions

0.73

channelAvailability

0.71

Activations Density 0.128%

phrases showing confusion or being confused

mentions of confusion and related states of uncertainty

No Known Activations