INDEX

Explanations

specific organizations, entities, and proper nouns mentioned in interviews or articles

oai_token-act-pair · gpt-3.5-turbo

names of organizations, companies, or countries

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

given

-0.65

*.

-0.59

.''.

-0.57

ppard

-0.54

 destro

-0.54

Yug

-0.54

*/(

-0.53

aceae

-0.53

.'"

-0.51

obyl

-0.51

POSITIVE LOGITS

Media

0.65

Leaks

0.64

 Observer

0.61

 Newsp

0.60

 Oversight

0.60

 Herald

0.59

Blog

0.59

news

0.57

Channel

0.56

Privacy

0.56

Activations Density 1.136%

specific organizations, entities, and proper nouns mentioned in interviews or articles

names of organizations, companies, or countries

No Comments

No Known Activations