INDEX

Explanations

the word "Perfect" in varying contexts and phrases

oai_token-act-pair · gpt-3.5-turbo

the repeated use of the word "Perfect" with varying intensities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

UGH

-0.71

UTH

-0.70

 veins

-0.69

 Sawyer

-0.66

 Cornell

-0.65

externalActionCode

-0.65

 epid

-0.63

ITNESS

-0.63

SIL

-0.63

 unpublished

-0.62

POSITIVE LOGITS

ion

1.01

entimes

0.87

ity

0.87

erion

0.86

imum

0.86

intendent

0.84

ly

0.84

nesses

0.82

ation

0.82

perfect

0.80

Activations Density 0.009%

the word "Perfect" in varying contexts and phrases

the repeated use of the word "Perfect" with varying intensities

No Comments

No Known Activations