INDEX

Explanations

the word "remains" in different contexts

oai_token-act-pair · gpt-3.5-turbo

the word "remains" indicating ongoing situations or states

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

ramid

-0.87

isson

-0.70

PsyNetMessage

-0.65

MAT

-0.65

idium

-0.65

DragonMagazine

-0.64

ongyang

-0.64

Lex

-0.62

 externalToEVAOnly

-0.62

 Globe

-0.62

POSITIVE LOGITS

 unchanged

1.27

 unanswered

1.24

 intact

1.23

 unsolved

1.18

 untouched

1.12

 unresolved

1.04

 undefeated

0.99

 steadfast

0.98

 unaffected

0.98

 undecided

0.97

Activations Density 0.037%

the word "remains" in different contexts

the word "remains" indicating ongoing situations or states

No Comments

No Known Activations