INDEX

Explanations

unique textual patterns within a specific format

oai_token-act-pair · gpt-3.5-turbo

special formatting or specific symbols typically associated with lists or data representation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

Ô

-0.98

ä»£

-0.78

theless

-0.74

 Avalon

-0.72

©¶æ

-0.70

acion

-0.68

ãģį

-0.67

heimer

-0.67

Interstitial

-0.66

ä¸Ń

-0.64

POSITIVE LOGITS

ep

0.78

roo

0.66

 better

0.64

cin

0.63

gap

0.62

EE

0.62

 Bill

0.61

Gov

0.61

ebus

0.61

geon

0.61

Activations Density 0.000%

unique textual patterns within a specific format

special formatting or specific symbols typically associated with lists or data representation

No Comments

No Known Activations