INDEX

Explanations

expressions of trust and mistrust towards people and entities

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

trust someone

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

 Bolo

-0.46

PyErr

-0.38

 occasioned

-0.38

riques

-0.37

writerow

-0.36

 additional

-0.36

lichkeiten

-0.36

 дополнитель

-0.36

Minor

-0.35

charCodeAt

-0.35

POSITIVE LOGITS

 trust

1.15

 Trust

1.10

Trust

1.07

 TRUST

1.06

trust

1.05

 Trusted

1.02

 trusting

1.02

 trusts

1.02

 trusted

0.97

trusted

0.94

Activations Density 0.009%

expressions of trust and mistrust towards people and entities

trust someone

No Comments

No Known Activations