INDEX

Explanations

mentions of physical pain

oai_token-act-pair · gpt-3.5-turbo

references to pain and related conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

indal

-0.85

eton

-0.83

sonian

-0.82

riminal

-0.82

ellar

-0.73

VERTISEMENT

-0.72

eor

-0.71

entric

-0.70

itures

-0.70

ordan

-0.69

POSITIVE LOGITS

staking

1.35

killers

1.28

killer

1.14

 pain

1.12

 inflicted

1.09

 agony

1.05

lessly

0.98

pain

0.95

 reliever

0.92

killing

0.90

Activations Density 0.031%

mentions of physical pain

references to pain and related conditions

No Comments

No Known Activations