INDEX

Explanations

words related to law enforcement operations or agencies

oai_token-act-pair · gpt-3.5-turbo

mentions of law enforcement or government agents

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

lihood

-0.82

ctory

-0.71

imeo

-0.68

issance

-0.68

fp

-0.68

STRUCT

-0.67

FORMATION

-0.66

nz

-0.65

FACE

-0.64

PUT

-0.64

POSITIVE LOGITS

 agents

1.48

 agent

1.30

agents

1.28

 Agents

1.23

agent

1.14

 Agent

1.04

ilage

0.92

 warr

0.85

Agent

0.84

 prov

0.82

Activations Density 0.012%

words related to law enforcement operations or agencies

mentions of law enforcement or government agents

No Comments

No Known Activations