INDEX

Explanations

terms related to interrogation activities

oai_token-act-pair · gpt-3.5-turbo

terms related to interrogation techniques

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

Offline

-0.76

erry

-0.74

lessness

-0.73

aim

-0.72

SY

-0.72

minecraft

-0.72

bc

-0.71

Merit

-0.68

 spare

-0.68

bearing

-0.67

POSITIVE LOGITS

 interrogation

1.36

 interrog

1.31

 interrogated

1.04

 Techniques

0.89

 quer

0.79

 torture

0.78

 techniques

0.76

isen

0.74

 questioning

0.74

 probing

0.73

Activations Density 0.014%

terms related to interrogation activities

terms related to interrogation techniques

No Comments

No Known Activations