INDEX

Explanations

tokens referring to previously mentioned subjects or concepts within the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 10-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_10/width_16k/average_l0_77

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.10.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

מו

-0.48

vespa

-0.47

trägen

-0.47

 eneste

-0.45

umsi

-0.44

待遇

-0.44

estal

-0.43

Es

-0.42

斗

-0.42

www

-0.42

POSITIVE LOGITS

 lenker

0.85

 كومونز

0.82

 BrowserModule

0.78

 referrerpolicy

0.77

AddHtmlAttribute

0.74

 MainAxisSize

0.73

ViewFeatures

0.73

ſelf

0.71

Efq

0.70

IBarButtonItem

0.69

Activations Density 0.266%

tokens referring to previously mentioned subjects or concepts within the text

No Comments

No Known Activations