INDEX

Explanations

words related to actions or processes happening in a specific context

oai_token-act-pair · gpt-3.5-turbo

terms related to processes and actions involving transition and cleanup efforts

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

BuyableInstoreAndOnline

-0.59

arij

-0.55

 Cosponsors

-0.53

ãĥ¼ãĥĨãĤ£

-0.52

 cameo

-0.51

Grab

-0.50

mber

-0.49

uploads

-0.49

 Nanto

-0.49

 typo

-0.48

POSITIVE LOGITS

 safely

1.19

 efficiently

1.15

 properly

1.03

 correctly

0.94

 effectively

0.91

 worthwhile

0.89

0.88

 reliably

0.87

orthy

0.85

 securely

0.84

Activations Density 0.587%

words related to actions or processes happening in a specific context

terms related to processes and actions involving transition and cleanup efforts

No Comments

No Known Activations