INDEX

Explanations

instances of significant increases in numeric or statistical trends

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

end of sentence

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

 geox

-0.49

VERTISEMENT

-0.49

principalColumn

-0.49

lapia

-0.49

FRAG

-0.47

Retries

-0.46

FED

-0.46

㋡

-0.45

CHIP

-0.45

edicated

-0.45

POSITIVE LOGITS

 disponibilités

0.49

تقاوى

0.44

 '\\;'

0.43

DockStyle

0.41

Skocz

0.40

 beginnetje

0.38

 Wiktionnaire

0.37

الدراسه

0.36

❋

0.35



0.35

Activations Density 0.393%

instances of significant increases in numeric or statistical trends

end of sentence

No Comments

No Known Activations