INDEX

Explanations

nouns related to problem-solving or discussions

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_fs6144-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_6144

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

6,144

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

inav

-0.66

mbuds

-0.62

OUP

-0.60

cluded

-0.57

iaries

-0.57

arnaev

-0.56

byter

-0.56

 Landing

-0.56

ADRA

-0.56

vae

-0.55

POSITIVE LOGITS

iest

1.10

 anew

0.96

of

0.89

liest

0.81

osphere

0.77

washer

0.76

 itself

0.75

 thereof

0.74

ifice

0.73

posts

0.73

Activations Density 0.937%

nouns related to problem-solving or discussions

No Comments

No Known Activations