INDEX

Explanations

proper nouns related to entertainment, such as movies, TV shows, and musicals

oai_token-act-pair · gpt-3.5-turbo

instances of the verb "to be" in various tenses

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ERAL

-0.74

ints

-0.73

pite

-0.72

 Cosponsors

-0.70

know

-0.69

arms

-0.68

root

-0.68

abuse

-0.67

address

-0.66

upper

-0.66

POSITIVE LOGITS

 narrated

1.35

 marketed

1.34

 premiered

1.31

 released

1.27

 filmed

1.24

 rated

1.23

 produced

1.20

 featured

1.18

 slated

1.18

 titled

1.15

Activations Density 0.207%

proper nouns related to entertainment, such as movies, TV shows, and musicals

instances of the verb "to be" in various tenses

No Comments

No Known Activations