INDEX

Explanations

words and phrases related to coding and programming errors

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 1-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_1/width_16k/average_l0_69

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

SequentialGroup

-0.51

featureID

-0.51

EndContext

-0.48

 Männern

-0.47

 purpoſe

-0.46

Inst

-0.45

writeFieldEnd

-0.44

@"/

-0.44

 témoignage

-0.43

 sumpay

-0.43

POSITIVE LOGITS

ным

0.96

них

0.88

ными

0.88

ческим

0.85

ные

0.85

ческих

0.84

чних

0.84

ных

0.84

ální

0.84

чным

0.83

Activations Density 0.279%

words and phrases related to coding and programming errors

No Comments

No Known Activations