INDEX

Explanations

references to training and educational programs

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

training and related contexts

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

VersionUID

-0.58

RRect

-0.53

boca

-0.51

 Ewig

-0.49

EDEFAULT

-0.49

<bos>

-0.48

 InputDecoration

-0.47

esch

-0.47

 Bellow

-0.47

bez

-0.46

POSITIVE LOGITS

 training

1.05

 Training

1.01

Training

0.96

 TRAINING

0.92

training

0.90

TRAINING

0.88

 trainings

0.84

 train

0.77

 treinamento

0.76

 Train

0.75

Activations Density 0.015%

references to training and educational programs

training and related contexts

No Comments

No Known Activations