INDEX

Explanations

words relating to people's names or titles

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

ãĤ¼ãĤ¦ãĤ¹

-0.72

Spread

-0.65

 loophole

-0.63

SPONSORED

-0.60

 conveniently

-0.58

\<

-0.57

 needle

-0.56

 matched

-0.55

 contradictions

-0.55

SEE

-0.52

POSITIVE LOGITS

kefeller

0.99

issance

0.82

zl

0.82

restling

0.79

hani

0.74

backer

0.74

bery

0.73

earchers

0.72

ophone

0.71

undo

0.69

Activations Density 0.075%

words relating to people's names or titles

No Comments

No Known Activations