INDEX

Explanations

phrases related to interviews or statements done by individuals

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

âĸº

-0.68

BLE

-0.67

root

-0.66

 Defin

-0.64

YES

-0.61

stones

-0.60

BOOK

-0.60

 spouses

-0.60

scope

-0.59

verting

-0.59

POSITIVE LOGITS

abolic

1.14

agram

1.02

oint

1.01

ointed

0.97

cients

0.95

alyses

0.91

 hour

0.91

 attempt

0.90

onym

0.90

aer

0.89

Activations Density 0.235%

phrases related to interviews or statements done by individuals

No Comments

No Known Activations