INDEX

Explanations

political titles and affiliations of individuals

oai_token-act-pair · gpt-3.5-turbo

references to political party affiliations and leadership roles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

gor

-0.65

 proxies

-0.63

 ponies

-0.62

tics

-0.61

-0.60

Synopsis

-0.60

 compilation

-0.59

 prol

-0.57

 prompt

-0.56

 CODE

-0.54

POSITIVE LOGITS

roit

0.71

jri

0.69

.,

0.69

iable

0.68

 Emer

0.65

ollo

0.64

 Pasadena

0.63

 Luxembourg

0.63

iffe

0.63

 Managing

0.63

Activations Density 0.100%

political titles and affiliations of individuals

references to political party affiliations and leadership roles

No Comments

No Known Activations