INDEX

Explanations

words related to sports and events, specifically related to football and fan gatherings

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

instances of media or content attribution and licensing

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

Ĥª

-0.80

backer

-0.68

gow

-0.65

ographers

-0.64

ava

-0.64

oms

-0.61

anooga

-0.60

\\\\\\\\

-0.59

 circulation

-0.59

rophic

-0.59

POSITIVE LOGITS

 sanity

0.67

ãĤ´

0.64

enable

0.63

 Style

0.61

ABLE

0.59

Ct

0.59

team

0.59

allowed

0.58

 dearly

0.58

ãĥ¯

0.58

Activations Density 0.231%

words related to sports and events, specifically related to football and fan gatherings

instances of media or content attribution and licensing

No Comments

No Known Activations