INDEX

Explanations

references to officials and their statements in a governmental context

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

officials speaking or reporting

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

 pinulongan

-0.57

 transfieras

-0.53

httphttps

-0.49

jsonwebtoken

-0.43

 فريبيس

-0.42

 bumble

-0.42

Tikang

-0.42

aarrggbb

-0.41

Portale

-0.41

WriteTagHelper

-0.41

POSITIVE LOGITS

 explained

0.58

explained

0.56

 announced

0.55

+#+#

0.51

 Officials

0.50

Officials

0.50

 officials

0.50

 briefed

0.47

 confirmed

0.47

 explicó

0.46

Activations Density 0.007%

references to officials and their statements in a governmental context

officials speaking or reporting

No Comments

No Known Activations