INDEX

Explanations

names of individuals with varying levels of importance

oai_token-act-pair · gpt-3.5-turbo

proper names, such as individuals and their titles or affiliations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

ModLoader

-0.58

 envy

-0.57

âĶĢâĶĢ

-0.57

 ultras

-0.56

 inheritance

-0.56

 underwater

-0.55

 subsistence

-0.55

 charity

-0.55

 deduction

-0.54

 Rebirth

-0.54

POSITIVE LOGITS

ansky

1.09

inski

1.02

etti

0.97

atz

0.97

opoulos

0.96

owski

0.95

zinski

0.94

gaard

0.94

lett

0.93

meyer

0.93

Activations Density 0.480%

names of individuals with varying levels of importance

proper names, such as individuals and their titles or affiliations

No Comments

No Known Activations