INDEX

Explanations

references to the word "Mad" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

chnology

-0.82

ģ«

-0.80

Ĥİ

-0.77

lished

-0.73

ç«

-0.71

hene

-0.70

ãģ¯

-0.69

VEL

-0.68

çĦ

-0.68

akeru

-0.67

POSITIVE LOGITS

agascar

1.56

onna

1.05

ness

1.01

rid

0.98

olyn

0.83

ame

0.83

apter

0.83

oka

0.82

olini

0.81

eline

0.80

Activations Density 0.004%

references to the word "Mad" in various contexts

No Comments

No Known Activations