INDEX

Explanations

phrases related to essential questions or statements that reveal significant insights

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

multi-cultural or real nightmare

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

uxxxx

-0.59

LookAnd

-0.53

 <>",

-0.51

 aapt

-0.47

RTGC

-0.46

 Exactos

-0.40

Jîn

-0.39

CodeAttribute

-0.38

 Periodicals

-0.38

henswürdigkeiten

-0.38

POSITIVE LOGITS

oggetto

0.43

:///

0.42

edar

0.41

раздо

0.40

TagMode

0.40

httphttps

0.39

ulman

0.39

zen

0.39

šė

0.39

inge

0.38

Activations Density 0.047%

phrases related to essential questions or statements that reveal significant insights

multi-cultural or real nightmare

No Comments

No Known Activations