INDEX

Explanations

phrases introducing additional information or related points

oai_token-act-pair · gpt-3.5-turbo

the word "Also" indicating additional or supplementary information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

acious

-0.81

agus

-0.80

erers

-0.69

acers

-0.66

adoes

-0.66

ouses

-0.64

uous

-0.63

ILLE

-0.62

 demolition

-0.61

aces

-0.61

POSITIVE LOGITS

 Known

0.83

 Included

0.83

 noteworthy

0.78

epad

0.76

 Called

0.71

 Quote

0.71

 includes

0.70

onday

0.70

 includ

0.69

 Notes

0.66

Activations Density 0.020%

phrases introducing additional information or related points

the word "Also" indicating additional or supplementary information

No Comments

No Known Activations