INDEX

Explanations

references to specific publications or works in scholarly 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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 destro

-0.77

 imperson

-0.74

 racing

-0.71

 spitting

-0.69

 occas

-0.68

 fart

-0.68

 riding

-0.67

 stomp

-0.67

 driving

-0.67

 enthusi

-0.66

POSITIVE LOGITS

âĵĺ

1.36

doi

1.06

McC

1.01

âĨ

0.99

Abstract

0.97

 PubMed

0.95

doi

0.93

References

0.92

http

0.91

 Horowitz

0.91

Activations Density 0.037%

references to specific publications or works in scholarly contexts

No Comments

No Known Activations