INDEX

Explanations

mentions of adversaries or foes

oai_token-act-pair · gpt-3.5-turbo

references to adversarial characters or entities

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

lic

-0.83

eret

-0.79

otide

-0.77

auntlets

-0.74

ced

-0.73

cer

-0.70

otto

-0.70

oled

-0.69

Shot

-0.69

UTH

-0.69

POSITIVE LOGITS

 enemies

1.24

foe

1.15

 enemy

1.12

 adversaries

1.11

 Enemies

1.07

 foes

1.05

 Enemy

0.99

 undermin

0.96

 adversary

0.91

emies

0.89

Activations Density 0.010%

mentions of adversaries or foes

references to adversarial characters or entities

No Comments

No Known Activations