INDEX

Explanations

references to methods of execution

oai_token-act-pair · gpt-3.5-turbo

mentions of execution and related terms

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

ijn

-0.86

ĨĴ

-0.72

thora

-0.72

MAG

-0.69

Tour

-0.69

ï¸ı

-0.67

 Seller

-0.66

Fi

-0.65

asio

-0.65

 Discord

-0.64

POSITIVE LOGITS

 executions

1.16

 execution

1.13

 executed

1.00

 execute

0.98

 inmate

0.91

 executes

0.86

 sentence

0.85

 penalty

0.83

 executing

0.82

 inmates

0.82

Activations Density 0.058%

references to methods of execution

mentions of execution and related terms

No Comments

No Known Activations