INDEX

Explanations

words indicating quality or effectiveness

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 25-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_25/width_16k/average_l0_116

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.25.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

WebVitals

-0.58

IUrlHelper

-0.58

 gehalten

-0.56

 aveug

-0.55

 fometimes

-0.54

 ſeveral

-0.53

 Foire

-0.53

 soigneusement

-0.53

coccal

-0.52

cocci

-0.52

POSITIVE LOGITS

 grounded

0.57

 intention

0.55

Portail

0.54

 known

0.54

 thought

0.53

 meaning

0.53

ografija

0.53

 versed

0.52

 liked

0.51

grounded

0.51

Activations Density 0.149%

words indicating quality or effectiveness

No Comments

No Known Activations