INDEX

Explanations

terms related to taxonomy and classification

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

enfance

-0.40

allclose

-0.36

 Chwiliwch

-0.36

 celebrid

-0.36

 étrangers

-0.35

 enfans

-0.35

 AssemblyCulture

-0.35

pyplot

-0.35

ulante

-0.34

 Romains

-0.34

POSITIVE LOGITS

tvguidetime

0.68

MLLoader

0.54

 GenerationType

0.53

ConstraintMaker

0.53

RegressionTest

0.50

 kohdetta

0.49

WireFormatLite

0.49

UnsafeEnabled

0.48

WriteTagHelper

0.47

AndEndTag

0.46

Activations Density 1.069%

terms related to taxonomy and classification

No Comments

No Known Activations