INDEX

Explanations

pronouns related to possession and reference

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

 understatement

-0.63

 Polk

-0.63

 Kelley

-0.58

 Father

-0.55

roads

-0.54

 quir

-0.53

Iowa

-0.53

 Cars

-0.53

 Herrera

-0.53

 mockery

-0.52

POSITIVE LOGITS

 accordingly

1.11

alian

1.00

atic

0.94

anwhile

0.90

ividual

0.78

atically

0.78

unes

0.76

rification

0.75

 appropriately

0.73

 afterwards

0.72

Activations Density 0.146%

pronouns related to possession and reference

No Comments

No Known Activations