INDEX

Explanations

words related to injuries or harm, with a focus on casualties

oai_token-act-pair · gpt-3.5-turbo

terms related to casualties and fatalities

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

perm

-0.80

Gos

-0.67

ramid

-0.66

efer

-0.65

cer

-0.65

hist

-0.65

aired

-0.64

ribed

-0.64

aton

-0.63

nder

-0.62

POSITIVE LOGITS

 casualties

1.36

 casualty

1.14

 bystanders

0.90

 losses

0.82

 fatalities

0.80

oided

0.80

 Victims

0.77

 inflicted

0.77

 deaths

0.75

 victims

0.75

Activations Density 0.008%

words related to injuries or harm, with a focus on casualties

terms related to casualties and fatalities

No Comments

No Known Activations