INDEX

Explanations

numbers and specific data points such as scores, years, and statistics

oai_token-act-pair · gpt-3.5-turbo

numbers, particularly years or counts

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

andise

-0.74

reen

-0.69

aye

-0.68

aeus

-0.68

acebook

-0.67

eton

-0.67

achu

-0.66

eners

-0.65

ende

-0.65

atoon

-0.65

POSITIVE LOGITS

0.87

0.86

0.77

0.75

²¾

0.75

0.72

0.71

0.70

MHz

0.69

 DEBUG

0.67

Activations Density 0.066%

numbers and specific data points such as scores, years, and statistics

numbers, particularly years or counts

No Comments

No Known Activations