INDEX

Explanations

expressions of personal recommendation or endorsement

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 39-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_39/width_16k/average_l0_131

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.39.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

Proposals

-0.72

 Proposals

-0.66

 proposals

-0.66

 suggestions

-0.63

 Vorschläge

-0.62

 Suggestions

-0.61

proposals

-0.60

 proposal

-0.59

Proposal

-0.59

suggestions

-0.57

POSITIVE LOGITS

 highly

1.20

highly

0.99

 Highly

0.96

Highly

0.88

 HIGH

0.82

 strongly

0.81

 altamente

0.77

 reco

0.71

 heartily

0.69

 recommend

0.69

Activations Density 0.195%

expressions of personal recommendation or endorsement

No Comments

No Known Activations