INDEX

Explanations

mentions of interrogation-related words and phrases

oai_token-act-pair · gpt-3.5-turbo

terms related to interrogation methods and practices

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

Offline

-0.79

ensical

-0.73

jri

-0.71

minecraft

-0.70

yright

-0.69

buy

-0.68

ouf

-0.67

erry

-0.66

WE

-0.65

cakes

-0.65

POSITIVE LOGITS

 interrogation

1.01

 interrog

0.99

 Techniques

0.89

 interrogated

0.85

 techniques

0.84

 questioning

0.68

 probing

0.68

 sessions

0.67

atories

0.67

 coercive

0.67

Activations Density 0.018%

mentions of interrogation-related words and phrases

terms related to interrogation methods and practices

No Comments

No Known Activations