INDEX

Explanations

phrases related to formal investigations

oai_token-act-pair · gpt-3.5-turbo

phrases indicating an investigation or inquiry

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

TA

-0.75

chart

-0.68

eret

-0.66

nown

-0.66

 ancest

-0.66

nick

-0.66

hop

-0.65

anguage

-0.65

reference

-0.65

oire

-0.64

POSITIVE LOGITS

 whether

0.87

 probing

0.83

 Integrity

0.80

 UFOs

0.73

 Scientology

0.72

 Probe

0.70

 wrongdoing

0.69

 Investigation

0.68

 allegations

0.67

Minimum

0.66

Activations Density 0.033%

phrases related to formal investigations

phrases indicating an investigation or inquiry

No Comments

No Known Activations