INDEX

Explanations

conjunctions and the word "and" in various contexts

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

ãĤª

-0.68

ocative

-0.65

anza

-0.63

urd

-0.63

agame

-0.63

payer

-0.62

criminal

-0.61

enance

-0.60

isi

-0.59

FILE

-0.58

POSITIVE LOGITS

 wondered

1.06

 luckily

0.97

 noticed

0.92

romeda

0.92

 THEN

0.91

 then

0.88

 waited

0.87

ersen

0.86

saw

0.86

 fortunately

0.84

Activations Density 0.147%

conjunctions and the word "and" in various contexts

No Comments

No Known Activations