INDEX

Explanations

negative statements or phrases with a 'not' followed by a verb or adjective

oai_token-act-pair · gpt-3.5-turbo

phrases that include the word "not" in various contexts

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

 Rebellion

-0.66

çĦ

-0.63

assembly

-0.63

IAS

-0.62

çĶŁ

-0.62

¿½

-0.61

 Mayhem

-0.61

LET

-0.60

Trivia

-0.59

ãĥĨ

-0.58

POSITIVE LOGITS

 necessarily

0.96

 realize

0.93

 agree

0.88

 realise

0.87

be

0.86

 qualify

0.83

necess

0.80

 survive

0.79

 have

0.77

 approve

0.77

Activations Density 0.057%

negative statements or phrases with a 'not' followed by a verb or adjective

phrases that include the word "not" in various contexts

No Comments

No Known Activations