INDEX

Explanations

technical terminology related to data processing and machine learning methodologies

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

machine learning methods classifiers

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

protoimpl

-0.44

rungsseite

-0.43

aarrggbb

-0.42

rowser

-0.41

LNB

-0.40

tolua

-0.39

 erat

-0.39

 kvinder

-0.39

 enfans

-0.39

⊝

-0.38

POSITIVE LOGITS

FormTagHelper

0.53

 Савезне

0.46

SequentialGroup

0.44

$_['

0.42

SupportActionBar

0.41

InputTagHelper

0.40

évaluateur

0.40

 Wikiseite

0.40

exactly

0.40

RegressionTest

0.39

Activations Density 0.050%

technical terminology related to data processing and machine learning methodologies

machine learning methods classifiers

No Comments

No Known Activations