INDEX

Explanations

terms and concepts related to timing, specifically the words "early" and "late" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_0/width_16k/average_l0_105

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.0.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

raman

-0.48

aka

-0.46

ngths

-0.46

kep

-0.45

 maximal

-0.45

huizen

-0.45

лежа

-0.44

 arreglar

-0.44

zcz

-0.43

 دنیا

-0.43

POSITIVE LOGITS

 early

1.09

early

1.06

 مشين

0.94

Early

0.85

 EARLY

0.83

EARLY

0.83

 late

0.81

+#+#

0.78

 Seed

0.78

Seed

0.78

Activations Density 0.175%

terms and concepts related to timing, specifically the words "early" and "late" in various contexts

No Comments

No Known Activations