INDEX

Explanations

phrases contrasting two different perspectives or pieces of information

oai_token-act-pair · gpt-3.5-turbo

contrasting ideas or perspectives

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

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

semb

-0.66

asley

-0.64

-0.63

STAR

-0.63

 Saras

-0.62

 Annotations

-0.61

stein

-0.61

 suffice

-0.59

before

-0.59

icago

-0.58

POSITIVE LOGITS

 srfAttach

0.77

 opposite

0.73

)].

0.65

cul

0.65

Õ

0.64

 second

0.63

 latter

0.62

ouple

0.61

middle

0.60

cum

0.60

Activations Density 0.060%

phrases contrasting two different perspectives or pieces of information

contrasting ideas or perspectives

No Comments

No Known Activations