INDEX

Explanations

numerical values mentioned as part of a statement

oai_token-act-pair · gpt-3.5-turbo

numerical expressions related to quantities and numbers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

TextColor

-0.74

EStream

-0.74

netflix

-0.73

ques

-0.68

Reviewer

-0.67

rely

-0.66

unct

-0.66

 tradem

-0.64

tainment

-0.64

atem

-0.62

POSITIVE LOGITS

%"

1.02

acity

0.95

th

0.93

0.88

0.87

 percent

0.86

0.85

Activations Density 0.052%

numerical values mentioned as part of a statement

numerical expressions related to quantities and numbers

No Comments

No Known Activations