INDEX

Explanations

specific markers or control instructions in a data structure or code format

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 7-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_7/width_16k/average_l0_69

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.7.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

 Gerv

-0.74

Symp

-0.67

 Trotz

-0.66

 Maurice

-0.65

Veh

-0.64

pertoire

-0.63

 قار

-0.62

Veh

-0.61

mace

-0.61

oston

-0.60

POSITIVE LOGITS

édrale

0.98

bleau

0.91

addGap

0.85

windowFixed

0.81

0.78



0.78

Diwedd

0.77

NavController

0.77

tometer

0.76

arakhand

0.75

Activations Density 0.134%

specific markers or control instructions in a data structure or code format

No Comments

No Known Activations