INDEX

Explanations

phrases indicating doubt or possibility

oai_token-act-pair · gpt-3.5-turbo

phrases that suggest uncertainty or speculation

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

lins

-0.78

waters

-0.75

raint

-0.74

board

-0.73

eries

-0.72

elson

-0.71

rix

-0.70

kowski

-0.70

raged

-0.69

ioch

-0.69

POSITIVE LOGITS

 misunder

0.87

querque

0.84

 jeopard

0.80

interstitial

0.79

 forgiven

0.75

merce

0.74

 infer

0.73

uthor

0.72

 surv

0.71

 swayed

0.70

Activations Density 0.017%

phrases indicating doubt or possibility

phrases that suggest uncertainty or speculation

No Comments

No Known Activations