INDEX

Explanations

references to specific items or entities being considered important or recommended

oai_token-act-pair · gpt-3.5-turbo

recurring references to movies and their significance

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

 undertake

-0.74

 periodic

-0.64

 foregoing

-0.64

 unden

-0.64

ustomed

-0.63

 Â§Â§

-0.62

Explore

-0.62

farious

-0.62

 observe

-0.62

 Parables

-0.61

POSITIVE LOGITS

 wasn

1.43

 sucks

1.40

 sucked

1.40

was

1.38

WAS

1.36

 sounded

1.35

 blew

1.34

 lasted

1.32

 didnt

1.29

 smelled

1.27

Activations Density 0.533%

references to specific items or entities being considered important or recommended

recurring references to movies and their significance

No Comments

No Known Activations