INDEX

Explanations

terms related to physical injuries

oai_token-act-pair · gpt-3.5-turbo

mentions of injuries

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

zees

-0.87

zee

-0.77

 Helpful

-0.70

esters

-0.68

zsche

-0.67

bara

-0.65

theless

-0.65

avanaugh

-0.65

zer

-0.64

ebus

-0.63

POSITIVE LOGITS

 injury

0.97

 inflicted

0.92

 injuries

0.89

 incurred

0.84

 inflic

0.83

oided

0.81

 wors

0.77

 injure

0.75

 prevention

0.74

juries

0.72

Activations Density 0.016%

terms related to physical injuries

mentions of injuries

No Comments

No Known Activations