INDEX

Explanations

references to ambiguous or unspecified concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_4/width_16k/average_l0_124

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.4.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

ঃ

-0.80

Datuak

-0.79

prefixer

-0.79

UpInside

-0.76

:^(

-0.74

 spalle

-0.74

titian

-0.71

 Kays

-0.71

 Reihen

-0.70

ietf

-0.69

POSITIVE LOGITS

 something

2.62

something

2.52

Something

2.38

 Something

2.34

 SOMETHING

2.07

 somethin

1.89

 thing

1.56

ETHING

1.55

thing

1.52

 anything

1.52

Activations Density 0.081%

references to ambiguous or unspecified concepts

No Comments

No Known Activations

references to ambiguous or unspecified concepts

No Comments

No Known Activations