INDEX

Explanations

phrases related to policies and ideals

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

amaz

-0.73

kernel

-0.66

otos

-0.61

ModLoader

-0.61

termin

-0.61

 mattress

-0.60

izont

-0.60

orman

-0.59

Spoiler

-0.59

toggle

-0.58

POSITIVE LOGITS

 alike

1.59

 thereof

1.11

 respectively

1.03

 relating

0.91

 belonging

0.86

 therein

0.85

 pertaining

0.84

gal

0.80

hips

0.80

 depending

0.79

Activations Density 0.258%

phrases related to policies and ideals

No Comments

No Known Activations