INDEX

Explanations

proper nouns or named entities

oai_token-act-pair · gpt-3.5-turbo

phrases that denote the beginning of statements or observations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ded

-0.46

abl

-0.45

edin

-0.43

sylv

-0.43

Ec

-0.43

 seed

-0.42

arsity

-0.42

tab

-0.42

 lasts

-0.42

oided

-0.42

POSITIVE LOGITS

etheless

0.96

odore

0.77

romeda

0.75

ccording

0.73

resa

0.72

theless

0.71

nsic

0.67

withstanding

0.67

swers

0.65

terness

0.64

Activations Density 0.659%

proper nouns or named entities

phrases that denote the beginning of statements or observations

No Comments

No Known Activations