INDEX

Explanations

phrases related to initiating processes or events

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

start/begin sequences

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

-0.31

Related

-0.27

 trade

-0.27

-0.26

://

-0.26

Trade

-0.26

tagHelperRunner

-0.26

</td>

-0.24

ậy

-0.24

POSITIVE LOGITS

ðsíða

0.77

 Starts

0.73

 starts

0.71

 Geiſt

0.71



0.71

majánló

0.71

iſchen

0.71

beginning

0.70

<unused16>

0.70

 beginning

0.70

Activations Density 0.028%

phrases related to initiating processes or events

start/begin sequences

No Comments

No Known Activations