INDEX

Explanations

words related to dice or gambling

oai_token-act-pair · gpt-3.5-turbo

references to dice

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

SUP

-0.74

ppard

-0.72

country

-0.72

ded

-0.72

nee

-0.71

organ

-0.71

onyms

-0.69

bed

-0.67

REM

-0.66

nant

-0.65

POSITIVE LOGITS

 dice

1.15

 Dice

1.14

 Royale

0.75

eria

0.74

 pool

0.74

 rolls

0.73

 Cutter

0.71

 gamble

0.69

acebook

0.68

 Dise

0.66

Activations Density 0.006%

words related to dice or gambling

references to dice

No Known Activations