INDEX

Explanations

names of places or countries

oai_token-act-pair · gpt-3.5-turbo

mentions of specific individuals, especially names

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

ĵĺ

-0.87

wcs

-0.59

CLASSIFIED

-0.56

 Adren

-0.55

 Corpus

-0.54

 seiz

-0.53

mx

-0.53

Enlarge

-0.51

 âĢº

-0.51

tags

-0.49

POSITIVE LOGITS

rite

0.55

onga

0.55

erno

0.55

orage

0.53

rafted

0.53

osuke

0.52

illet

0.51

idon

0.51

eto

0.51

 ports

0.49

Activations Density 6.879%

names of places or countries

mentions of specific individuals, especially names

No Comments

No Known Activations