INDEX

Explanations

mentions of attire or clothing

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "bes," indicating a focus on representation or status

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

Not in Any Lists

No Comments

Negative Logits

 Marijuana

-0.68

 Examiner

-0.66

 District

-0.65

 Kubrick

-0.65

 Records

-0.64

osaurs

-0.62

 Lilly

-0.62

 Reds

-0.62

 Oval

-0.61

 Roosevelt

-0.61

POSITIVE LOGITS

 challeng

1.08

bes

0.96

entimes

0.95

poke

0.87

iege

0.84

sembly

0.84

semb

0.83

icker

0.82

aved

0.81

bridge

0.80

Activations Density 0.006%

mentions of attire or clothing

occurrences of the word "bes," indicating a focus on representation or status

No Comments

No Known Activations