INDEX

Explanations

pairs of words indicating options or alternatives

oai_token-act-pair · gpt-3.5-turbo

phrases indicating the presence of alternative options or choices

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

ONSORED

-0.70

 Haku

-0.70

ADVERTISEMENT

-0.65

ECH

-0.65

 Dragonbound

-0.63

 Reporter

-0.63

ETS

-0.62

 Photographer

-0.61

 Candidate

-0.61

 Miko

-0.61

POSITIVE LOGITS

chard

1.33

nam

1.08

ific

1.07

ifice

1.07

chid

1.07

acle

1.07

acular

1.04

nery

0.99

acles

0.97

lando

0.95

Activations Density 0.078%

pairs of words indicating options or alternatives

phrases indicating the presence of alternative options or choices

No Comments

No Known Activations