INDEX

Explanations

weapons and violent actions

oai_token-act-pair · gpt-3.5-turbo

terms associated with violence and conflict

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

natureconservancy

-0.66

ĻĤ

-0.65

algia

-0.64

 arrang

-0.63

 confidentiality

-0.62

 Salary

-0.60

 Sanctuary

-0.59

Priv

-0.58

ffen

-0.58

 Differences

-0.55

POSITIVE LOGITS

 onto

1.51

 into

1.42

 toward

1.16

 towards

1.13

 overboard

1.11

 INTO

1.06

into

1.03

 darts

0.97

 projectiles

0.96

 Into

0.96

Activations Density 0.285%

weapons and violent actions

terms associated with violence and conflict

No Comments

No Known Activations