INDEX

Explanations

terms related to instructions and guidance, particularly in a technical or computational context

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

instruction, Instructs, Instron

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.31.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

乓

-0.38

subpackage

-0.38

 üzere

-0.38

TagMode

-0.38

postmedia

-0.35

étape

-0.35

arrêt

-0.34

 Warga

-0.32

annonce

-0.31

enderror

-0.31

POSITIVE LOGITS

 nahilalakip

0.66

 Instr

0.60

+#+#

0.57

Instr

0.54

 eiffel

0.53

Дереккөздер

0.51

cyclopedia

0.51

 ſind

0.50

évaluateur

0.50

 insur

0.49

Activations Density 0.021%

terms related to instructions and guidance, particularly in a technical or computational context

instruction, Instructs, Instron

No Comments

No Known Activations