INDEX

Explanations

statements related to allegations, discoveries, and confirmations of events or issues

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

click

-0.67

opter

-0.62

itas

-0.58

<<

-0.55

||

-0.54

agu

-0.54

adena

-0.53

turn

-0.53

Tonight

-0.52

Sem

-0.52

POSITIVE LOGITS

autions

0.68

iosyncr

0.68

 anecd

0.66

interstitial

0.65

ractive

0.63

saf

0.62

racted

0.61

nesses

0.60

andals

0.59

 endorsements

0.59

Activations Density 0.357%

statements related to allegations, discoveries, and confirmations of events or issues

No Comments

No Known Activations