INDEX

Explanations

direct or indirect statements made by various individuals

oai_token-act-pair · gpt-3.5-turbo

phrases that refer to comments, statements, or interactions among individuals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

ETHOD

-0.67

oiler

-0.65

 inver

-0.62

properties

-0.62

 tert

-0.61

tics

-0.61

thumbnails

-0.61

mobi

-0.61

enegger

-0.60

inately

-0.59

POSITIVE LOGITS

 bestowed

1.11

 handed

0.96

 undertaken

0.93

 levied

0.91

 emanating

0.86

 exchanged

0.84

 issued

0.82

 uttered

0.82

oled

0.80

 taken

0.79

Activations Density 0.921%

direct or indirect statements made by various individuals

phrases that refer to comments, statements, or interactions among individuals

No Comments

No Known Activations