INDEX

Explanations

code-related constructs and terminology from programming languages

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

C++ const ref templates

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.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

+#+#

-0.43

ISupport

-0.41

 chip

-0.41

 finish

-0.39

WithIOException

-0.38

 đủ

-0.37

 Aliens

-0.37

仗

-0.36

 rock

-0.35

 Alien

-0.35

POSITIVE LOGITS

Geplaatst

0.63

StoryboardSegue

0.59

 AssemblyVersion

0.52

InitVars

0.49

isContained

0.48

Filmographie

0.47

 resourceCulture

0.46

原始内容

0.46

таратура

0.46

 nakalista

0.46

Activations Density 0.015%

code-related constructs and terminology from programming languages

C++ const ref templates

No Comments

No Known Activations