INDEX

Explanations

dimensions and reshaping functions related to array manipulation

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

unique indices and shape manipulation

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

 understatement

-0.40

fwd

-0.39

ⓘ

-0.38

PMailer

-0.38

┍

-0.36

 Goes

-0.35

ambu

-0.35

Throwaway

-0.35

 courants

-0.35

POSTS

-0.35

POSITIVE LOGITS

Jeografia

0.57

GraphicsUnit

0.56

PerformLayout

0.52

RegressionTest

0.48

IVEREF

0.47

IUrlHelper

0.44

StructEnd

0.44

 olympique

0.44

UnusedPrivate

0.44

 Ravi

0.43

Activations Density 0.002%

dimensions and reshaping functions related to array manipulation

unique indices and shape manipulation

No Comments

No Known Activations