INDEX

Explanations

numerical data or representations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 15-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_15/width_16k/average_l0_78

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.15.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

 themſelves

-0.81

 CreateTagHelper

-0.81

 ſtill

-0.76

 AssemblyCulture

-0.74

 fubject

-0.73

 juſt

-0.71

 raiſ

-0.69

 reaſon

-0.68

 ſever

-0.68

ftagPool

-0.67

POSITIVE LOGITS

expandindo

0.55

 Roskov

0.55

üyor

0.50

El

0.48

wife

0.48

 bodies

0.47

Sh

0.45

bular

0.45

ielo

0.45

elve

0.43

Activations Density 0.351%

numerical data or representations

No Comments

No Known Activations