INDEX

Explanations

actions indicating a process or a necessary step

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 12-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_12/width_16k/average_l0_130

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.12.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

RegressionTest

-0.56

utives

-0.50

 surla

-0.49

 EconPapers

-0.48

Tē

-0.48

 Catawiki

-0.47

UserScript

-0.46

 queſta

-0.43

cshtml

-0.43

RTGC

-0.42

POSITIVE LOGITS

 must

0.82

 Must

0.81

Must

0.78

must

0.75

 MUST

0.67

MUST

0.64

 harus

0.63

 Gotta

0.61

 devons

0.61

 gotta

0.61

Activations Density 1.192%

actions indicating a process or a necessary step

No Comments

No Known Activations