INDEX

Explanations

occurrences of the word "Cole" in the text

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

oppable

-0.69

 recogn

-0.66

 Mehran

-0.61

channelAvailability

-0.60

CRIP

-0.58

oslav

-0.58

ĺħ

-0.58

 Munich

-0.58

 loader

-0.58

hots

-0.57

POSITIVE LOGITS

tto

1.17

tti

1.02

voy

0.90

tta

0.88

mons

0.86

opter

0.86

bourne

0.83

brate

0.82

craft

0.81

gon

0.79

Activations Density 0.028%

occurrences of the word "Cole" in the text

No Comments

No Known Activations