INDEX

Explanations

references to the United States

oai_token-act-pair · gpt-3.5-turbo

references to the United States

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 Noir

-0.77

Redditor

-0.73

 Preferred

-0.72

*/(

-0.72

 Cologne

-0.70

 Siren

-0.66

CDs

-0.65

 Waiting

-0.63

 Featured

-0.62

 Pioneer

-0.62

POSITIVE LOGITS

prising

1.04

nexpected

0.96

seless

0.92

raviolet

0.91

NA

0.87

pperc

0.84

mpire

0.84

UU

0.81

lyss

0.81

olt

0.81

Activations Density 0.049%

references to the United States

references to the United States

No Comments

No Known Activations