INDEX

Explanations

mentions of specific names or surnames, especially when followed by numbers

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of a particular name or term associated with people or entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 glim

-0.75

Subject

-0.73

IDENT

-0.73

glass

-0.71

 narrator

-0.70

Þ

-0.69

 à¨

-0.68

icist

-0.66

IAL

-0.66

 iPads

-0.66

POSITIVE LOGITS

zzi

1.26

ppa

1.06

zzo

1.03

pps

1.03

ichi

0.96

eta

0.93

opa

0.92

ÅĤ

0.92

ek

0.91

ven

0.91

Activations Density 0.008%

mentions of specific names or surnames, especially when followed by numbers

repeated mentions of a particular name or term associated with people or entities

No Comments

No Known Activations