INDEX

Explanations

occurrences where discussions lead to a solution or resolution

oai_token-act-pair · gpt-3.5-turbo

end-of-document markers

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

 Jagu

-0.69

evidence

-0.68

 Vaugh

-0.68

Edit

-0.64

chuk

-0.60

Links

-0.60

Events

-0.57

Contents

-0.57

Hon

-0.57

Ont

-0.57

POSITIVE LOGITS

lot

1.04

 bunch

1.01

 handful

0.89

 plethora

0.89

 huge

0.88

rouse

0.88

 couple

0.86

few

0.85

 glimpse

0.82

 piece

0.82

Activations Density 0.869%

occurrences where discussions lead to a solution or resolution

end-of-document markers

No Comments

No Known Activations