INDEX

Explanations

terms related to military conflicts and nuclear deterrence

oai_token-act-pair · gpt-3.5-turbo

references to military conflicts and threats

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 Cosponsors

-0.86

 Parables

-0.77

 Lounge

-0.77

 Journals

-0.74

 Loft

-0.73

NB

-0.73

 anecd

-0.72

 crochet

-0.72

 Volunte

-0.71

POSITIVE LOGITS

 invading

1.36

 aggress

1.34

 milit

1.34

 invaders

1.33

 invasion

1.33

 enemy

1.29

 adversary

1.29

 adversaries

1.27

 conquest

1.26

 Soviets

1.26

Activations Density 0.645%

terms related to military conflicts and nuclear deterrence

references to military conflicts and threats

No Comments

No Known Activations