INDEX

Explanations

mentions of the word "odo" followed by a number (e.g., "odo 10")

oai_token-act-pair · gpt-3.5-turbo

mentions of the word "odo" and associated context within a specific framework

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

Negative Logits

lon

-0.76

rail

-0.74

retion

-0.74

ting

-0.71

par

-0.68

rac

-0.68

 Stras

-0.67

ipl

-0.67

riz

-0.67

pet

-0.67

POSITIVE LOGITS

odo

0.87

CRIP

0.83

herty

0.76

éĹĺ

0.71

 Ahead

0.69

Ã±a

0.69

externalActionCode

0.68

ãģ®å

0.68

aird

0.66

DoS

0.65

Activations Density 0.017%

mentions of the word "odo" followed by a number (e.g., "odo 10")

mentions of the word "odo" and associated context within a specific framework

No Known Activations