INDEX

Explanations

people's names

oai_token-act-pair · gpt-3.5-turbo

occurrences of specific keywords and proper nouns

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 Matth

-0.83

NCT

-0.73

 Assassin

-0.72

hei

-0.71

================

-0.70

 Cumm

-0.70

ogly

-0.69

cuc

-0.69

 Breat

-0.68

Cig

-0.68

POSITIVE LOGITS

urdue

0.95

ung

0.86

roy

0.81

ten

0.80

water

0.77

wald

0.76

orean

0.76

ACTED

0.76

bara

0.76

uned

0.76

Activations Density 0.427%

people's names

occurrences of specific keywords and proper nouns

No Comments

No Known Activations