INDEX

Explanations

specific items or entities mentioned in a list

oai_token-act-pair · gpt-3.5-turbo

references to social and political issues involving marginalization or exclusion

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ãĥį

-0.85

exclusive

-0.69

\/\/

-0.65

Beg

-0.63

ym

-0.60

ãĥ©ãĥ³

-0.60

 HuffPost

-0.60

ãĥ«

-0.60

Ö¼

-0.60

 Piper

-0.59

POSITIVE LOGITS

etc

1.62

 blah

1.08

âĢ¦)

1.07

etc

1.02

...)

1.02

âĢ¦

0.95

,...

0.89

 âĢ¦

0.88

âĢ¦.

0.85

ect

0.83

Activations Density 0.399%

specific items or entities mentioned in a list

references to social and political issues involving marginalization or exclusion

No Comments

No Known Activations