INDEX

Explanations

concepts related to moral or ethical judgments involving rewards and punishments

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

drawbacks, negatives, or downsides

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.31.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

tagext

-0.41

UpInside

-0.35

 eikä

-0.35

PickerController

-0.34

inigte

-0.32

whitelist

-0.32

◓

-0.32

 Himmel

-0.32

 Excessive

-0.31

Fernsehserie

-0.31

POSITIVE LOGITS

 drawbacks

0.74

 detriment

0.73

 negatives

0.72

 detrimental

0.70

 detri

0.66

 downsides

0.66

 negativa

0.65

 disappointments

0.65

 downside

0.65

 conversely

0.65

Activations Density 0.178%

concepts related to moral or ethical judgments involving rewards and punishments

drawbacks, negatives, or downsides

No Comments

No Known Activations