INDEX

Explanations

phrases indicating familiarity or knowledge with a subject

oai_token-act-pair · gpt-3.5-turbo

references to familiarity or knowledge of concepts, entities, or experiences

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

ishers

-0.74

cers

-0.72

eters

-0.68

gencies

-0.68

itter

-0.68

aredevil

-0.68

isher

-0.65

isco

-0.63

anmar

-0.63

pex

-0.63

POSITIVE LOGITS

lege

1.04

 firsthand

0.91

ledge

0.90

 anecd

0.79

 intimately

0.73

 existed

0.69

 exists

0.68

terday

0.65

 instinctively

0.63

âĦ¢:

0.63

Activations Density 0.056%

phrases indicating familiarity or knowledge with a subject

references to familiarity or knowledge of concepts, entities, or experiences

No Comments

No Known Activations