INDEX

Explanations

phrases expressing subjective evaluations or opinions about various topics

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

mycin

-0.34

 Torrance

-0.33

')}}

-0.32

wahati

-0.32

 portata

-0.32

phrine

-0.31

ături

-0.30

WillAppear

-0.30

asanjo

-0.30

DIM

-0.29

POSITIVE LOGITS

Ideally

0.59

 Ideally

0.59

本当は

0.58

AddTagHelper

0.57

findpost

0.57

期刊论文

0.55

featureID

0.54

 ModelRenderer

0.54

يكب

0.52

AsUp

0.51

Activations Density 0.726%

phrases expressing subjective evaluations or opinions about various topics

No Comments

No Known Activations