INDEX

Explanations

phrases related to potential harm or danger to individuals or society

oai_token-act-pair · gpt-3.5-turbo

concerns related to the impact of actions on public health and safety

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

irection

-0.65

 brisk

-0.64

 boast

-0.61

 caveats

-0.60

 bluff

-0.59

 authoritative

-0.59

 manoeuv

-0.55

 flurry

-0.55

 motions

-0.54

 suggestive

-0.54

POSITIVE LOGITS

 financially

0.94

 unborn

0.86

ankind

0.85

 downstream

0.83

 wellbeing

0.83

 economically

0.82

 elector

0.81

 harmed

0.81

 livelihood

0.79

 outwe

0.79

Activations Density 0.380%

phrases related to potential harm or danger to individuals or society

concerns related to the impact of actions on public health and safety

No Comments

No Known Activations