INDEX

Explanations

names or proper nouns, particularly related to historical or mythical figures

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Configuration

neuronpedia/gpt2-small__res_scl-ajt/6-res_scl-ajt

Prompts (Dashboard)

12,288 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

46,080

Data Type

torch.float32

Hook Point

blocks.6.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

apollo-research/Skylion007-openwebtext-tokenizer-gpt2

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

rency

-1.06

ãĥ¼ãĥĨ

-0.94

fitting

-0.90

enment

-0.84

pmwiki

-0.81

ãĥł

-0.81

BILITIES

-0.80

FML

-0.79

ãĥĥãĥī

-0.78

ovember

-0.77

POSITIVE LOGITS

oslav

1.11

ovych

1.04

akov

0.97

abytes

0.94

Yug

0.86

doi

0.85

gi

0.82

aston

0.82

gm

0.79

 VIDEOS

0.78

Activations Density 15.923%

names or proper nouns, particularly related to historical or mythical figures

No Comments

No Known Activations