INDEX

Explanations

descriptions related to visual appearance and perception

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_21/width_16k/average_l0_70

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

 subjectivity

-0.56

 كومونز

-0.56

FormTagHelper

-0.55

ParallelGroup

-0.54

Jegyzetek

-0.51

 obscurity

-0.51

 dinilai

-0.51

 kijken

-0.50

ServiceModel

-0.48

 migli

-0.48

POSITIVE LOGITS

 like

1.02

 promising

0.64

 good

0.57

EDEFAULT

0.56

 exactly

0.55

fjspx

0.55

 familiar

0.54

 promis

0.53

 likes

0.53

DockStyle

0.53

Activations Density 0.082%

descriptions related to visual appearance and perception

No Comments

No Known Activations