INDEX

Explanations

occurences of the word "text" and its variations in the input

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

text

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

showConfirm

-0.45

hasMany

-0.45

 rantai

-0.45

augmenter

-0.45

 Alliance

-0.44

@"/

-0.44

 Osorio

-0.44

作为一个

-0.43

 CHAMPION

-0.43

 bencana

-0.42

POSITIVE LOGITS

 text

0.94

 Text

0.86

Text

0.85

TEXT

0.77

 getText

0.75

 texts

0.73

 TEXT

0.73

getText

0.73

text

0.73

 texto

0.72

Activations Density 0.025%

occurences of the word "text" and its variations in the input

text

No Comments

No Known Activations