INDEX

Explanations

expressions of trust and emotional reactions to situations

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

why or when followed by explanation

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

principalColumn

-0.39

 jsPsych

-0.36

 stationnement

-0.33

 nocturn

-0.33

kháu

-0.32

osv

-0.32

.*")]

-0.32

 monasterio

-0.31

 colab

-0.31

 cerve

-0.31

POSITIVE LOGITS

Personensuche

0.65

IsMutable

0.60

endphp

0.58

 initComponents

0.57

 Normdatei

0.57

RTHOOK

0.55

 resourceCulture

0.52

rungsseite

0.50

ValueStyle

0.50

DoubleQuotes

0.50

Activations Density 0.049%

expressions of trust and emotional reactions to situations

why or when followed by explanation

No Comments

No Known Activations