INDEX

Explanations

mentions of the Drug Enforcement Administration and related terms

oai_token-act-pair · gpt-3.5-turbo

references to the DEA and drug-related terminology

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Forge

-0.79

#$

-0.77

lass

-0.73

issance

-0.71

pread

-0.71

Ã¸

-0.70

elong

-0.69

faces

-0.68

hello

-0.67

bird

-0.66

POSITIVE LOGITS

 cartel

1.08

DEA

1.07

 narcotics

1.06

 cartels

1.03

 traffickers

0.98

 trafficking

0.98

 narc

0.91

 Traff

0.88

 forfeiture

0.87

 seizure

0.86

Activations Density 0.030%

mentions of the Drug Enforcement Administration and related terms

references to the DEA and drug-related terminology

No Comments

No Known Activations