INDEX

Explanations

occurrences of the word "and."

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

nor

-0.70

èĢħ

-0.68

 Weekly

-0.68

Week

-0.66

etheless

-0.66

ZA

-0.65

æŃ¦

-0.65

é¾

-0.65

ancial

-0.63

 Juliet

-0.62

POSITIVE LOGITS

 downs

0.66

 down

0.64

yss

0.63

aunder

0.63

 scratch

0.61

orthy

0.60

 grabs

0.60

xual

0.60

oots

0.60

 saddle

0.59

Activations Density 0.035%

occurrences of the word "and."

No Comments

No Known Activations