INDEX

Explanations

references to mental and psychological states or conditions

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

psychological warfare

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.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

 gynhyrchwyd

-0.58

 trám

-0.55

HostException

-0.55

árbol

-0.54

erçe

-0.54

流动

-0.52

 Suppliers

-0.51

 dostaw

-0.50

 Buses

-0.50

farwydd

-0.50

POSITIVE LOGITS

 psychological

0.66

Psychological

0.63

 psychologically

0.59

 mental

0.57

 Psychological

0.56

心理

0.54

Mental

0.51

 psyche

0.50

psychological

0.49

 psicológico

0.49

Activations Density 0.027%

references to mental and psychological states or conditions

psychological warfare

No Comments

No Known Activations