INDEX

Explanations

mention of proper nouns and usernames

oai_token-act-pair · gpt-3.5-turbo

specific identifiers or names

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

soType

-0.78

 confir

-0.75

 dilig

-0.71

 contrace

-0.70

INF

-0.68

 psychiat

-0.66

 indecent

-0.66

 antic

-0.65

 Instr

-0.65

 electronically

-0.65

POSITIVE LOGITS

1.16

1.04

kowski

1.01

kb

1.01

kan

1.00

ked

0.98

KC

0.98

ks

0.97

KEN

0.95

kok

0.94

Activations Density 0.174%

mention of proper nouns and usernames

specific identifiers or names

No Comments

No Known Activations