INDEX

Explanations

film-related details such as movie titles, film directors, and actor names

oai_token-act-pair · gpt-3.5-turbo

references to movies and films

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

tu

-0.69

 unanim

-0.66

 waivers

-0.66

respons

-0.65

 handwriting

-0.62

Standing

-0.62

 exemption

-0.60

ollah

-0.60

compliance

-0.59

 somet

-0.59

POSITIVE LOGITS

 premiered

1.51

 debuted

1.37

 starred

1.31

 premie

1.26

 revolves

1.18

 stars

1.15

 centers

1.15

 follows

1.15

 depicts

1.11

 explores

1.11

Activations Density 0.174%

film-related details such as movie titles, film directors, and actor names

references to movies and films

No Comments

No Known Activations