INDEX

Explanations

references to watches

oai_token-act-pair · gpt-3.5-turbo

mentions of various types of watches

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

Reviewer

-0.92

«ĺ

-0.77

 congr

-0.72

Interstitial

-0.71

istically

-0.70

EngineDebug

-0.68

emia

-0.65

å¥³

-0.65

bably

-0.64

 chemotherapy

-0.64

POSITIVE LOGITS

tower

1.47

dogs

1.21

dog

1.19

watch

0.99

points

0.93

strap

0.91

list

0.90

point

0.89

maker

0.88

fulness

0.87

Activations Density 0.046%

references to watches

mentions of various types of watches

No Comments

No Known Activations