INDEX

Explanations

phrases related to harmful or dangerous situations

oai_token-act-pair · gpt-3.5-turbo

phrases and terms related to toxic substances and environmental hazards

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

Negative Logits

ramids

-0.70

 specialization

-0.70

Trend

-0.69

odore

-0.68

ilater

-0.68

boxing

-0.68

thood

-0.67

 downgrade

-0.66

Role

-0.66

Requirements

-0.65

POSITIVE LOGITS

 liquids

1.60

 fluids

1.60

 saliva

1.49

 liquid

1.49

 fluid

1.48

 gases

1.42

 saline

1.38

 urine

1.36

 molten

1.34

 fumes

1.34

Activations Density 0.528%

phrases related to harmful or dangerous situations

phrases and terms related to toxic substances and environmental hazards

No Known Activations