INDEX

Explanations

the word "Derby" at different activations

oai_token-act-pair · gpt-3.5-turbo

mentions of Derby and individuals related to it

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

itar

-0.95

ities

-0.93

ITIES

-0.90

itÃ©

-0.87

itary

-0.86

ropri

-0.86

ified

-0.84

uador

-0.84

itol

-0.81

itatively

-0.80

POSITIVE LOGITS

lli

0.86

 Devils

0.78

 detector

0.74

Fax

0.73

DoS

0.73

lla

0.71

 Derby

0.69

 sting

0.68

 Gibbs

0.67

 detectors

0.67

Activations Density 0.042%

the word "Derby" at different activations

mentions of Derby and individuals related to it

No Comments

No Known Activations