INDEX

Explanations

mentions of lobes or lobotomy

oai_token-act-pair · gpt-3.5-turbo

references to anatomical features and terms related to lobes and specific individuals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Hunger

-0.80

idon

-0.74

ordan

-0.73

wyn

-0.71

ostics

-0.71

Wet

-0.71

 ILCS

-0.71

 Loll

-0.68

resist

-0.67

 Jelly

-0.67

POSITIVE LOGITS

lob

2.44

Dob

1.44

 spectacular

1.42

 countryside

1.26

 lobe

1.22

 Nice

1.20

nikov

1.12

Nice

1.06

Lob

1.02

 lobster

1.01

Activations Density 0.036%

mentions of lobes or lobotomy

references to anatomical features and terms related to lobes and specific individuals

No Comments

No Known Activations