INDEX

Explanations

programming concepts related to data management and handling symbol operations

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

inserting or processing records

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

 فريبيس

-0.48

 greateſt

-0.44

 Theſe

-0.44

 ftate

-0.42

 becauſe

-0.41

 houſe

-0.40

LMA

-0.40

 zijne

-0.39

neſs

-0.39

 raiſ

-0.38

POSITIVE LOGITS

FDRE

0.45

 registered

0.45

newName

0.43

Personensuche

0.43

 honor

0.42

 Honor

0.42

ENODEV

0.41

dAtA

0.40

lineTo

0.40

 Brennan

0.38

Activations Density 0.019%

programming concepts related to data management and handling symbol operations

inserting or processing records

No Comments

No Known Activations