INDEX

Explanations

concepts related to blame and attribution of responsibility

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

 INTERESAR

-0.57

odori

-0.50

chemise

-0.48

 prototyping

-0.48

:✨

-0.47

pushFollow

-0.47

 Cárdenas

-0.45

履

-0.45

Mə

-0.45

Matty

-0.45

POSITIVE LOGITS

 blame

0.75

 blamed

0.69

 blaming

0.68

 causes

0.60

 blames

0.58

 attributed

0.56

 cause

0.55

 attribution

0.53

blame

0.51

 causas

0.49

Activations Density 1.011%

concepts related to blame and attribution of responsibility

No Comments

No Known Activations