INDEX

Explanations

phrases related to actions and instructions, particularly in a software or application context

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

confirmation and continuation choices

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.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

rungsseite

-0.66

 fieldNum

-0.56

errHandler

-0.55

 समीक्षाओं

-0.54

Portale

-0.54

 arşivlendi

-0.53

 nakalista

-0.53

 surla

-0.52

localctx

-0.52

verwijspagina

-0.52

POSITIVE LOGITS

Eventually

0.36

 Release

0.36

PositiveButton

0.34

 Eventually

0.32

blauch

0.32

VIAF

0.31

inheritDoc

0.31

Nevermind

0.30

 tani

0.30

 delay

0.29

Activations Density 0.010%

phrases related to actions and instructions, particularly in a software or application context

confirmation and continuation choices

No Comments

No Known Activations