INDEX

Explanations

instances where something is discovered or found through investigation

oai_token-act-pair · gpt-3.5-turbo

phrases related to discoveries and findings in investigative contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

externalActionCode

-0.79

anwhile

-0.76

Cooldown

-0.72

é¾įå¥ĳå£«

-0.72

âĨ

-0.71

ignore

-0.69

EStreamFrame

-0.69

adan

-0.67

 CrossRef

-0.67

 Entreprene

-0.67

POSITIVE LOGITS

 traces

1.06

 probable

0.96

 evidence

0.95

 irregularities

0.92

 fingerprints

0.86

 remnants

0.85

DNA

0.83

 incrim

0.83

 inconsistencies

0.81

 signs

0.80

Activations Density 0.155%

instances where something is discovered or found through investigation

phrases related to discoveries and findings in investigative contexts

No Comments

No Known Activations