INDEX

Explanations

specific names and titles related to companies, products, or services

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

 Entered

-0.82

 Gutenberg

-0.58

 Hutchinson

-0.57

Lub

-0.57

 spared

-0.56

 awoken

-0.54

yright

-0.53

icter

-0.53

 agitated

-0.53

ecause

-0.52

POSITIVE LOGITS

 respectively

0.73

 alike

0.70

rio

0.65

 itself

0.65

rect

0.64

ancel

0.64

wide

0.63

ÃĽ

0.62

lite

0.60

 Presents

0.59

Activations Density 0.206%

specific names and titles related to companies, products, or services

No Comments

No Known Activations