INDEX

Explanations

locations and actors involved in acts of violence or conflict

oai_token-act-pair · gpt-3.5-turbo

geographic locations and entities associated with conflict and terrorism

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Wide

-0.62

âĶĢâĶĢâĶĢâĶĢ

-0.62

tera

-0.61

FINE

-0.61

ificent

-0.60

ermanent

-0.60

ä½¿

-0.60

DEV

-0.60

Standard

-0.60

 Alphabet

-0.59

POSITIVE LOGITS

 targets

0.87

 intern

0.83

 unarmed

0.75

 invaders

0.72

 unjust

0.68

 protestors

0.68

dur

0.66

 unprepared

0.66

 tyres

0.66

 temples

0.65

Activations Density 0.623%

locations and actors involved in acts of violence or conflict

geographic locations and entities associated with conflict and terrorism

No Comments

No Known Activations