INDEX

Explanations

terms related to licensing and copyright information

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

Tokens followed by license or initials

np_acts-logits-general · gemini-2.0-flash

license and prefixes like BPF_

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_16k/average_l0_76

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.31.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

oa̍t

-0.35

 gehabt

-0.34

дента

-0.33

aronder

-0.32

WithMany

-0.31

 putus

-0.30

不然

-0.30

ppia

-0.29

 equally

-0.29

zugehen

-0.29

POSITIVE LOGITS

transQ

0.65

LLocation

0.60

BZ

0.55

BB

0.54

BM

0.53

følgelig

0.53

quelize

0.53

BH

0.52

BPD

0.52

BN

0.50

Activations Density 0.793%

terms related to licensing and copyright information

Tokens followed by license or initials

license and prefixes like BPF_

No Comments

No Known Activations