INDEX

Explanations

patterns related to structured data formats or programming syntax

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

code and math symbols

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

-0.40

-0.37

所

-0.36

-0.35

-0.34

-0.33

POSITIVE LOGITS

 فريبيس

0.90

 AssemblyCompany

0.80

WriteTagHelper

0.79

 виправивши

0.79

__::

0.79

*/),

0.77

 """",

0.77

 >=",

0.77

 AssemblyVersion

0.76

 tartalomajánló

0.74

Activations Density 0.089%

patterns related to structured data formats or programming syntax

code and math symbols

No Comments

No Known Activations