INDEX

Explanations

terms related to video games, tactics, and strategies in competitive gaming

oai_token-act-pair · gpt-3.5-turbo 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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

iannopoulos

-0.64

çĦ

-0.63

uez

-0.62

awa

-0.61

————————

-0.61

atics

-0.59

 reel

-0.59

owship

-0.59

hips

-0.58

ittal

-0.58

POSITIVE LOGITS

measures

1.11

fe

1.04

intuitive

1.02

parts

1.02

part

0.96

terrorism

0.95

balance

0.95

point

0.93

intelligence

0.93

attack

0.93

Activations Density 0.013%

terms related to video games, tactics, and strategies in competitive gaming

No Comments

No Known Activations