INDEX

Explanations

words related to lethal substances or actions

oai_token-act-pair · gpt-3.5-turbo

references to lethal situations or substances

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

acular

-0.77

arest

-0.77

agall

-0.76

ramid

-0.75

angles

-0.74

hew

-0.71

roo

-0.69

eon

-0.69

 Revised

-0.69

annis

-0.68

POSITIVE LOGITS

 injection

1.09

 lethal

1.03

 dose

1.01

 doses

0.86

icide

0.85

lethal

0.85

 injections

0.84

ously

0.84

espie

0.84

otine

0.83

Activations Density 0.009%

words related to lethal substances or actions

references to lethal situations or substances

No Comments

No Known Activations