INDEX

Explanations

references to the speaker or writer's engagement and actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

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

netflix

-0.74

cum

-0.68

nesota

-0.66

Appearances

-0.65

Reviewer

-0.62

idium

-0.61

atory

-0.60

 totality

-0.59

Ago

-0.58

case

-0.58

POSITIVE LOGITS

've

1.30

're

1.27

bsite

1.18

'll

1.14

 strive

1.06

 recommend

1.00

athered

0.96

'd

0.95

akening

0.94

 intend

0.94

Activations Density 0.193%

references to the speaker or writer's engagement and actions

No Comments

No Known Activations