INDEX

Explanations

phrases involving sequential actions or numbering

oai_token-act-pair · gpt-3.5-turbo

references to sequences or progressive actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

lees

-0.69

icals

-0.65

ables

-0.62

glers

-0.56

fields

-0.56

nc

-0.54

ustomed

-0.54

ancies

-0.54

Delta

-0.53

 deposits

-0.53

POSITIVE LOGITS

elfth

0.74

itans

0.73

 Confederation

0.69

©¶æ

0.69

enth

0.67

othy

0.66

 foremost

0.65

Two

0.65

 Atlantis

0.65

 equals

0.65

Activations Density 0.218%

phrases involving sequential actions or numbering

references to sequences or progressive actions

No Comments

No Known Activations