INDEX

Explanations

expressions related to falsehood or error validation in responses

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 11-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_11/width_16k/average_l0_118

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.11.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

 autorytatywna

-0.79

:✨

-0.74

 queſta

-0.58

rawDesc

-0.56

 wiſſen

-0.54

WaitGroup

-0.54

ekan

-0.51

awaiter

-0.51

 public

-0.50

⸅

-0.50

POSITIVE LOGITS

évaluateur

0.35

 vPvB

0.35

 yoksa

0.35

 vôtre

0.33

ckså

0.33

новништво

0.33

forhold

0.32

RenderAtEndOf

0.32

 virš

0.32

 yoktur

0.32

Activations Density 1.926%

expressions related to falsehood or error validation in responses

No Comments

No Known Activations