INDEX

Explanations

law-related terms and actions

oai_token-act-pair · gpt-3.5-turbo

specific legal or investigative terms

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

mington

-0.71

olded

-0.70

sworth

-0.69

 entrenched

-0.68

urden

-0.65

 Sturgeon

-0.63

jun

-0.62

meter

-0.60

 loaded

-0.60

stroke

-0.60

POSITIVE LOGITS

ĳå£«

0.86

IONS

0.77

IENT

0.75

ATIONS

0.75

ations

0.68

REDACTED

0.66

aea

0.66

 moderation

0.65

MENTS

0.64

ROR

0.64

Activations Density 0.214%

law-related terms and actions

specific legal or investigative terms

No Comments

No Known Activations