INDEX

Explanations

specific rules or guidelines mentioned in a text

oai_token-act-pair · gpt-3.5-turbo

mentions of rules and regulations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

ience

-0.82

ONSORED

-0.76

 Bras

-0.72

ienced

-0.65

 Astro

-0.63

Dak

-0.62

igated

-0.62

issance

-0.61

aged

-0.61

 experien

-0.61

POSITIVE LOGITS

 rules

0.99

 governing

0.96

Rule

0.93

 violations

0.90

Rules

0.88

book

0.87

breakers

0.82

DragonMagazine

0.82

 lawy

0.82

books

0.81

Activations Density 0.020%

specific rules or guidelines mentioned in a text

mentions of rules and regulations

No Comments

No Known Activations