INDEX

Explanations

words related to law enforcement and crime investigation

oai_token-act-pair · gpt-3.5-turbo

terms related to investigation and restrictions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 crane

-0.68

 DRAG

-0.62

 nerv

-0.62

 abundantly

-0.60

 deaf

-0.59

 foremost

-0.58

 ashore

-0.58

 crawl

-0.58

elsius

-0.58

REDACTED

-0.58

POSITIVE LOGITS

ments

1.77

ions

1.65

able

1.58

ables

1.49

ors

1.46

ibles

1.42

ible

1.36

ings

1.36

ability

1.34

ives

1.34

Activations Density 0.266%

words related to law enforcement and crime investigation

terms related to investigation and restrictions

No Comments

No Known Activations