INDEX

Explanations

statements about intensity or severity in narratives

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits



-0.55

UserScript

-0.54

 CallOverrides

-0.54

цездатний

-0.52

fortawesome

-0.51

 NSCoder

-0.50

ädie

-0.49

 IFTT

-0.48

 Anſ

-0.48

 вікі

-0.48

POSITIVE LOGITS

 high

0.41

 dangerous

0.39

 dangers

0.37

 caution

0.37

 hohen

0.36

 huge

0.36

 large

0.35

 tapete

0.35

 severe

0.34

 scary

0.34

Activations Density 0.087%

statements about intensity or severity in narratives

No Comments

No Known Activations