INDEX

Explanations

technical terms and programming-related phrases

oai_token-act-pair · gpt-3.5-turbo

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

imester

-0.62

umerous

-0.61

arer

-0.61

-0.60

Rated

-0.60

ISTORY

-0.60

sequently

-0.59

arest

-0.57

successfully

-0.56

Â©

-0.55

POSITIVE LOGITS

 weird

1.24

 stuff

1.20

 shit

1.16

 crap

1.15

 kinda

1.09

 shitty

1.07

 goddamn

1.06

 crappy

1.05

 fucking

1.04

 freaking

1.03

Activations Density 1.352%

technical terms and programming-related phrases

No Comments

No Known Activations