INDEX

Explanations

terms related to technical specifications and classifications in various contexts

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

after specific technical or numeric terms

np_acts-logits-general · gemini-2.0-flash

old ones

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 28-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_28/width_16k/average_l0_119

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.28.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

RegressionTest

-0.56

voltaic

-0.52

asitic

-0.51

orr

-0.50

idosis

-0.47

genodigd

-0.47

phite

-0.45

vician

-0.45

ofy

-0.45

nship

-0.45

POSITIVE LOGITS

 ones

1.13

のもの

0.70

なもの

0.68

するもの

0.68

したもの

0.67

 stuff

0.65

 hacerlo

0.63

もので

0.61

 варианты

0.59

 вариан

0.59

Activations Density 1.386%

terms related to technical specifications and classifications in various contexts

after specific technical or numeric terms

old ones

No Comments

No Known Activations