INDEX

Explanations

special characters or codes at the end of a phrase

oai_token-act-pair · gpt-3.5-turbo

negative indicators or terms related to issues or problems

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 proble

-0.66

 craving

-0.66

 shaping

-0.65

 endeav

-0.65

 reckoning

-0.65

 happ

-0.64

 outlet

-0.64

 leaps

-0.64

 recall

-0.64

 warming

-0.63

POSITIVE LOGITS

=-=-=-=-=-=-=-=-

1.24

=-=-=-=-

1.23

 Advertisement

1.06

_-

0.99

/-

0.90

Phoenix

0.89

 Stuff

0.88

Commercial

0.83

 Ibid

0.82

Detailed

0.81

Activations Density 0.042%

special characters or codes at the end of a phrase

negative indicators or terms related to issues or problems

No Comments

No Known Activations