INDEX

Explanations

elements related to claiming and requesting assistance or information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

MMS

-0.41

 Supra

-0.39

 Employers

-0.39

astify

-0.39

clés

-0.38

 Rites

-0.38

 aprons

-0.37

 diplomacy

-0.37

 proximal

-0.36

 daim

-0.36

POSITIVE LOGITS

 resourceCulture

0.57

principalTable

0.56

 entirely

0.55

 zupełnie

0.54

 altogether

0.52

 ogóle

0.51

 completely

0.50

 вовсе

0.50

 kokona

0.48

 вообще

0.47

Activations Density 0.720%

elements related to claiming and requesting assistance or information

No Comments

No Known Activations