INDEX

Explanations

phrases related to research conclusions or study results

oai_token-act-pair · gpt-3.5-turbo

punctuation marks, specifically commas

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

dict

-0.63

Mog

-0.62

oths

-0.60

 recommending

-0.60

 raids

-0.60

 Safari

-0.60

 Videos

-0.59

malink

-0.59

 favor

-0.59

-.

-0.59

POSITIVE LOGITS

æĺ¯

0.70

 Aust

0.70

itect

0.70

SEA

0.69

iddles

0.68

 folds

0.66

ipples

0.66

vic

0.66

onnaissance

0.65

Fram

0.64

Activations Density 0.000%

phrases related to research conclusions or study results

punctuation marks, specifically commas

No Comments

No Known Activations