INDEX

Explanations

names or titles indicating a prominent role such as lead author, founder, investigator, or creator

oai_token-act-pair · gpt-3.5-turbo

roles and titles of individuals in various contexts

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

Associated

-0.66

castle

-0.65

alus

-0.65

Supplement

-0.61

 cuff

-0.60

apest

-0.60

armac

-0.59

igslist

-0.59

solid

-0.59

POSITIVE LOGITS

 Richard

0.88

 Steph

0.86

Cao

0.84

 Sergey

0.84

 Mike

0.84

 Jeremy

0.83

 Derek

0.82

 Nigel

0.82

 Nicholas

0.81

 Ezek

0.81

Activations Density 0.272%

names or titles indicating a prominent role such as lead author, founder, investigator, or creator

roles and titles of individuals in various contexts

No Comments

No Known Activations