INDEX

Explanations

references to events captured on video

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

instances of events being documented on video or camera

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

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

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_post

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

bane

-0.72

%]

-0.69

ieties

-0.69

anders

-0.65

enance

-0.63

ocations

-0.62

rough

-0.61

çĶ°

-0.61

lore

-0.61

reads

-0.60

POSITIVE LOGITS

 CCTV

1.36

 television

1.21

 videot

1.19

 webcam

1.08

 satellite

1.07

 Youtube

1.07

TV

1.06

 camera

1.06

 tape

1.05

TVs

1.05

Activations Density 0.228%

references to events captured on video

instances of events being documented on video or camera

No Comments

No Known Activations