INDEX

Explanations

programming constructs and method calls in code snippets

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

programming variables and methods

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

expandindo

-0.86

 يتيمه

-0.77

__(/*!

-0.66

 pinulongan

-0.65

ViewInit

-0.64

nodoc

-0.64

OGND

-0.63

@"/

-0.62

principalTable

-0.62

setVerticalGroup

-0.61

POSITIVE LOGITS

final

0.33

is

0.32

ciaż

0.32

this

0.32

дца

0.31

 Dieser

0.30

we

0.30

 final

0.29

 дело

0.29

 item

0.29

Activations Density 0.034%

programming constructs and method calls in code snippets

programming variables and methods

No Comments

No Known Activations