INDEX

Explanations

contractions of "does not"

oai_token-act-pair · gpt-3.5-turbo

negations or forms of the word "doesn't."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

 palms

-0.62

ANG

-0.59

 Casting

-0.57

 embodiments

-0.57

âĹ¼

-0.56

 Amph

-0.56

 misdem

-0.56

 Butt

-0.55

armed

-0.55

 VERS

-0.54

POSITIVE LOGITS

't

1.42

etsk

0.93

kie

0.93

ĳå£«

0.89

berra

0.82

paces

0.81

tery

0.79

okia

0.79

ates

0.79

":"/

0.78

Activations Density 0.079%

contractions of "does not"

negations or forms of the word "doesn't."

No Comments

No Known Activations