INDEX

Explanations

proper nouns and specific entities mentioned in news articles

oai_token-act-pair · gpt-3.5-turbo

references to specific organizations, places, or notable entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

$.

-0.88

*.

-0.83

.</

-0.77

!.

-0.72

`.

-0.71

etc

-0.69

âĢ¦.

-0.68

."

-0.68

..

-0.68

.<

-0.68

POSITIVE LOGITS

 spokesman

0.99

 spokeswoman

0.95

 spokesperson

0.92

 Tribune

0.75

 historian

0.74

 memorandum

0.74

ologist

0.73

 Herald

0.71

 watchdog

0.70

 Gazette

0.70

Activations Density 0.793%

proper nouns and specific entities mentioned in news articles

references to specific organizations, places, or notable entities

No Comments

No Known Activations