INDEX

Explanations

locations described using block numbers

oai_token-act-pair · gpt-3.5-turbo

references to specific location blocks in an area

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

anson

-0.68

xon

-0.67

elsen

-0.66

inen

-0.66

emic

-0.66

atically

-0.65

 Trance

-0.63

 complication

-0.63

oleon

-0.62

earch

-0.62

POSITIVE LOGITS

busters

0.94

ages

0.87

quote

0.87

aded

0.86

aders

0.82

ades

0.78

chains

0.78

yards

0.75

yard

0.75

jer

0.73

Activations Density 0.017%

locations described using block numbers

references to specific location blocks in an area

No Comments

No Known Activations