INDEX

Explanations

references to various types of deaths or fatalities

oai_token-act-pair · gpt-3.5-turbo

mentions of fatalities or deaths, particularly in the context of violence or specific events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 Hilbert

-0.74

Cola

-0.73

PLA

-0.69

 Collider

-0.67

Stud

-0.65

ðĿ

-0.64

Alpha

-0.63

arity

-0.63

 Cube

-0.63

 Plex

-0.62

POSITIVE LOGITS

 deaths

1.10

 Deaths

0.91

umatic

0.83

death

0.75

soever

0.75

anasia

0.75

 reckoning

0.72

inations

0.72

 destro

0.72

 overdoses

0.72

Activations Density 0.011%

references to various types of deaths or fatalities

mentions of fatalities or deaths, particularly in the context of violence or specific events

No Comments

No Known Activations