INDEX

Explanations

blocks of text and their structure, particularly in code or documentation format

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-2b-pt-res/layer_1/width_16k/average_l0_102

Prompts (Dashboard)

36,864 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

SourceChecksum

-0.91

abestanden

-0.80

 nahilalakip

-0.80

__*/

-0.77



-0.71

+#+#

-0.70

UnusedPrivate

-0.68

expandindo

-0.68

 AssemblyCulture

-0.66

.*")]

-0.64

POSITIVE LOGITS

///

0.67

MLLoader

0.59

 Bourgoin

0.54

//

0.52

atche

0.49

fus

0.49

//!

0.48

enciaga

0.48

///

0.47

 uintptr

0.47

Activations Density 1.719%

blocks of text and their structure, particularly in code or documentation format

No Comments

No Known Activations