INDEX

Explanations

patterns related to negative and positive attributes in data formatting

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

negative numbers or symbols

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

mpions

-0.52

majánló

-0.50

ropriate

-0.49

idavit

-0.49

 utafitiHapana

-0.49

proportion

-0.49

 désolés

-0.47

TypedDataSet

-0.47

Proportion

-0.46

 proportionately

-0.46

POSITIVE LOGITS

Atoi

0.39

 noDo

0.37

TagMode

0.33

Παραπομπές

0.33

posedge

0.32

asaran

0.30

Források

0.30

setDo

0.30

 INTERESAR

0.30

Jegyzetek

0.29

Activations Density 0.014%

patterns related to negative and positive attributes in data formatting

negative numbers or symbols

No Comments

No Known Activations