INDEX

Explanations

keywords indicating importance or primary focus

oai_token-act-pair · gpt-3.5-turbo

words indicating primary or principal relationships among subjects or actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

Lyn

-0.77

anth

-0.75

ITH

-0.73

 Oaks

-0.70

 Mann

-0.68

 Chaser

-0.65

wal

-0.64

Afee

-0.64

ylan

-0.63

ilion

-0.62

POSITIVE LOGITS

 aimed

0.90

 geared

0.85

 focused

0.85

 consist

0.83

 relying

0.83

 concentrated

0.82

 reliant

0.82

tenance

0.81

 rely

0.78

xual

0.76

Activations Density 0.014%

keywords indicating importance or primary focus

words indicating primary or principal relationships among subjects or actions

No Comments

No Known Activations