INDEX

Explanations

past tense verbs ending in 'ed'

oai_token-act-pair · gpt-3.5-turbo

words that indicate past actions or states

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

å¥³

-0.77

itars

-0.70

perature

-0.69

 CLSID

-0.69

ORY

-0.68

perty

-0.68

abet

-0.68

å£«

-0.65

ascript

-0.65

udeb

-0.65

POSITIVE LOGITS

nesday

0.95

 aback

0.91

uled

0.81

own

0.80

ed

0.76

monton

0.75

adoes

0.73

uct

0.73

eals

0.72

ging

0.70

Activations Density 0.051%

past tense verbs ending in 'ed'

words that indicate past actions or states

No Comments

No Known Activations