INDEX

Explanations

phrases related to law enforcement or FBI agents

oai_token-act-pair · gpt-3.5-turbo

references to law enforcement agents

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

lehem

-0.86

orthern

-0.84

 proble

-0.75

ashington

-0.69

ashtra

-0.67

 pont

-0.67

 palace

-0.67

 finance

-0.67

 assembly

-0.67

wob

-0.66

POSITIVE LOGITS

 Agent

3.92

 Agents

3.04

Agent

2.68

agent

1.42

 Hitman

1.37

 agents

1.35

 agent

1.33

 Officer

1.30

 Trooper

1.29

 Detective

1.25

Activations Density 0.034%

phrases related to law enforcement or FBI agents

references to law enforcement agents

No Comments

No Known Activations