INDEX

Explanations

references to World War I and related military context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 12-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_12/width_16k/average_l0_130

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.12.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

 ModelExpression

-0.69

gjenge

-0.52

 arkiv

-0.52

setVerticalGroup

-0.52

tiérrez

-0.51

filé

-0.51

 zilver

-0.49

CloseHandle

-0.49

fclose

-0.48

 sorriso

-0.47

POSITIVE LOGITS

mobileqq

0.43

WWI

0.39

Abitanti

0.38

jsonwebtoken

0.37

unknownFields

0.35

bollah

0.35

Anhalt

0.34

lorette

0.34

ніципалі

0.33

obium

0.32

Activations Density 0.568%

references to World War I and related military context

No Comments

No Known Activations