INDEX

Explanations

phrases emphasizing correctness, accuracy, or confirmation

oai_token-act-pair · gpt-3.5-turbo

affirmations and confirmations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 agric

-0.68

 ropes

-0.67

tes

-0.66

plates

-0.65

 faults

-0.64

Pros

-0.64

 belts

-0.63

 selves

-0.60

assemb

-0.60

çīĪ

-0.60

POSITIVE LOGITS

SPONSORED

0.72

natureconservancy

0.69

 Mahjong

0.67

 copyrighted

0.65

 motivation

0.64

 scary

0.64

ummer

0.64

 Beir

0.62

aze

0.62

AppData

0.61

Activations Density 0.303%

phrases emphasizing correctness, accuracy, or confirmation

affirmations and confirmations

No Comments

No Known Activations