INDEX

Explanations

structured data elements, particularly those related to properties and formatting in a configuration or programming context

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

`{` and `:`

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

parsedMessage

-1.02

httphttps

-0.77

 ModelExpression

-0.76

ſicht

-0.75

 queſta

-0.75

AndEndTag

-0.71

WebElementEntity

-0.71

<pad>

-0.71

<unused14>

-0.70

[@BOS@]

-0.70

POSITIVE LOGITS

0.35

 sugg

0.31

↵↵

0.31

0.30

Consider

0.30

↵

0.30

0.29

IN

0.29

Activations Density 0.023%

structured data elements, particularly those related to properties and formatting in a configuration or programming context

`{` and `:`

No Comments

No Known Activations