INDEX

Explanations

phrases indicating information or instructions given to someone

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 ILCS

-0.67

vous

-0.64

prus

-0.63

 sidx

-0.62

aho

-0.62

adesh

-0.62

adish

-0.59

anu

-0.59

ockets

-0.59

aird

-0.59

POSITIVE LOGITS

by

1.22

 beforehand

0.96

 orally

0.93

 repeatedly

0.84

by

0.81

 aloud

0.77

tale

0.77

 aback

0.76

BY

0.75

 verbally

0.74

Activations Density 0.037%

phrases indicating information or instructions given to someone

No Comments

No Known Activations