INDEX

Explanations

references to names or titles, particularly those associated with notable individuals or events

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

worldly

-0.75

 RELEASE

-0.70

 omission

-0.68

merce

-0.65

EMBER

-0.64

ATIONAL

-0.62

 desirable

-0.62

 midterm

-0.61

 plunge

-0.60

 lockout

-0.60

POSITIVE LOGITS

enstein

1.29

ett

1.14

enberg

1.10

enh

1.10

enger

1.07

nell

1.04

erman

1.04

erton

1.01

etts

1.00

ens

0.98

Activations Density 0.018%

references to names or titles, particularly those associated with notable individuals or events

No Comments

No Known Activations