INDEX

Explanations

number-related details like dates and locations

oai_token-act-pair · gpt-3.5-turbo

sequences or formats related to numerical data and lists

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

enegger

-0.71

-0.70

 toast

-0.69

pend

-0.69

-0.68

 Peng

-0.67

heed

-0.66

 Chaser

-0.65

son

-0.65

POSITIVE LOGITS

0.85

0.84

ģ«

0.80

0.77

illes

0.76

²¾

0.72

wer

0.70

0.66

 Blackwell

0.65

uth

0.65

Activations Density 0.107%

number-related details like dates and locations

sequences or formats related to numerical data and lists

No Comments

No Known Activations