INDEX

Explanations

numbers corresponding to scoring in sports, specifically focusing on references to point values

oai_token-act-pair · gpt-3.5-turbo

references to scoring points in sports or games

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

 notor

-0.73

 underrated

-0.71

 unequ

-0.66

natureconservancy

-0.65

 medically

-0.63

 chemically

-0.63

 electronically

-0.62

 metabol

-0.61

 tradem

-0.61

ESE

-0.60

POSITIVE LOGITS

point

1.28

points

1.02

pointers

0.95

hole

0.90

patrick

0.86

eus

0.84

finding

0.81

gradient

0.81

ilib

0.80

edly

0.75

Activations Density 0.009%

numbers corresponding to scoring in sports, specifically focusing on references to point values

references to scoring points in sports or games

No Comments

No Known Activations