INDEX

Explanations

patterns related to numerical identifiers and codes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

 verre

-0.73

 smal

-0.69

Alban

-0.68

 Kongo

-0.67

 Fabio

-0.65

KD

-0.64

Wahl

-0.63

 المثال

-0.63

 Nava

-0.62

Irak

-0.62

POSITIVE LOGITS

cc

1.68

CC

1.60

cc

1.51

CC

1.43

CCB

1.12

Cc

1.05

Cc

1.04

CCS

0.98

 Chambers

0.97

RegressionTest

0.96

Activations Density 0.183%

patterns related to numerical identifiers and codes

No Comments

No Known Activations