INDEX

Explanations

references to conflicts, war, and international political issues

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

references to specific geopolitical entities and actors

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_post

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

 Hitch

-0.54

 corners

-0.52

 Cthulhu

-0.52

 Wilde

-0.52

Omn

-0.52

 monog

-0.51

 Chop

-0.51

 Sherlock

-0.50

zip

-0.50

 Lovecraft

-0.50

POSITIVE LOGITS

zbollah

1.00

Syria

0.88

addafi

0.88

merga

0.85

Syrian

0.81

xit

0.81

oshenko

0.80

Army

0.79

uterte

0.79

oros

0.75

Activations Density 0.200%

references to conflicts, war, and international political issues

references to specific geopolitical entities and actors

No Comments

No Known Activations