INDEX

Explanations

specific nouns related to political, intelligence, or military activities

oai_token-act-pair · gpt-3.5-turbo

references to individuals involved in intelligence and military operations

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

reath

-0.75

venge

-0.74

rounded

-0.73

ranged

-0.72

Bah

-0.70

grain

-0.69

reek

-0.69

gets

-0.69

alone

-0.68

Pr

-0.66

POSITIVE LOGITS

 operative

1.43

 operatives

1.03

eering

0.86

eers

0.75

 Emails

0.75

 tong

0.74

ãĤ¤ãĥĪ

0.73

 manuals

0.72

èĢħ

0.70

 mastermind

0.69

Activations Density 0.004%

specific nouns related to political, intelligence, or military activities

references to individuals involved in intelligence and military operations

No Comments

No Known Activations