INDEX

Explanations

phrases related to events and actions in various contexts

oai_token-act-pair · gpt-3.5-turbo

phrases and contexts related to investigations

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

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

à¦

-1.02

ãĥİ

-0.98

à

-0.92

âĢ

-0.87

 ç¥ŀ

-0.86

âķ

-0.86

 pione

-0.84

Your

-0.84

iPhone

-0.82

Picture

-0.81

POSITIVE LOGITS

 Rove

1.11

 Schwartz

1.06

 Wilkinson

1.02

 Franks

0.99

 Shields

0.98

 Rosenthal

0.98

 Bernstein

0.97

 Fitzgerald

0.96

 Perkins

0.96

 Schwarz

0.96

Activations Density 0.461%

phrases related to events and actions in various contexts

phrases and contexts related to investigations

No Comments

No Known Activations