INDEX

Explanations

references to conspiratorial or aggressive language targeting specific groups or individuals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 37-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_37/width_16k/average_l0_124

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.37.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

adaptiveStyles

-0.57

awtextra

-0.56

="@+

-0.51

odziel

-0.48

MarshalTo

-0.44

Descripció

-0.44

 Erwartungen

-0.43

 aventure

-0.42

DIPSETTING

-0.42

expandindo

-0.42

POSITIVE LOGITS

 fascist

0.85

 propaganda

0.84

 fucking

0.82

 leftist

0.79

 treason

0.78

 traitors

0.78

 fascism

0.77

 genoc

0.77

 patriots

0.75

 corrupt

0.75

Activations Density 1.800%

references to conspiratorial or aggressive language targeting specific groups or individuals

No Comments

No Known Activations