INDEX

Explanations

proper nouns, specifically names of individuals or entities

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ãĤ¼ãĤ¦ãĤ¹

-0.91

ãĥ´ãĤ¡

-0.73

ãĤ¦ãĤ¹

-0.66

 partName

-0.64

issance

-0.63

shown

-0.63

à¤¾

-0.62

metic

-0.62

 Janeiro

-0.61

INA

-0.60

POSITIVE LOGITS

zinski

0.70

acket

0.61

pson

0.61

 Samurai

0.59

bley

0.59

iott

0.59

otos

0.58

bye

0.58

kes

0.58

oller

0.57

Activations Density 0.032%

proper nouns, specifically names of individuals or entities

No Comments

No Known Activations