INDEX

Explanations

phrases or punctuation indicating emphasis or significance

oai_token-act-pair · gpt-3.5-turbo

instances of dashes or interruptions in text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 protective

-0.72

othes

-0.65

 packing

-0.65

ovi

-0.64

 Piece

-0.64

 oven

-0.63

ograph

-0.63

 basil

-0.62

 grave

-0.62

 cream

-0.61

POSITIVE LOGITS

lance

0.94

fuck

0.91

jobs

0.89

micro

0.83

something

0.81

sil

0.81

release

0.79

sat

0.79

like

0.79

assisted

0.78

Activations Density 0.014%

phrases or punctuation indicating emphasis or significance

instances of dashes or interruptions in text

No Comments

No Known Activations