INDEX

Explanations

terms related to electricity and electrostatics

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

 trekken

-0.47

 betreft

-0.38

Accordingly

-0.38

 Accordingly

-0.38

 valmis

-0.37

ALITY

-0.37

 Duarte

-0.36

格局

-0.35

 proactively

-0.35

GroupLayout

-0.34

POSITIVE LOGITS

WebVitals

0.68

adpleegd

0.65

AutoScaleMode

0.62

 experimento

0.59

@[+][

0.57

ỡng

0.56

 Numerade

0.56

makeText

0.54

WebControls

0.53

 mergeFrom

0.53

Activations Density 0.672%

terms related to electricity and electrostatics

No Comments

No Known Activations