© Neuronpedia 2026

Privacy & Terms Blog GitHub Slack Twitter Contact

Jacobian LensNEW

Natural Language

MAY 26 Assistant AxisJAN 26 Circuit TracerAPR 26Steer SAE Evals Exports interp-engineNEW Community Blog Privacy & Terms Contact

LLAMA3.1-8B-IT · 7-RESID-POST-AA · 21433 ｜ Neuronpedia

Home
Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
Llama3.1-8B-IT
Resid Post - 131k
7-RESID-POST-AA
21433

INDEX

Explanations

throne

np_max-act · gemini-2.0-flash

New Auto-Interp

Top Features by Cosine Similarity

Configuration

andyrdt/saes-llama-3.1-8b-instruct/resid_post_layer_7/trainer_1

Dataset (Dashboard)

Various

Features

131,072

Data Type

float32

Hook Name

blocks.7.hook_resid_post

Architecture

standard

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

Show PlotsShow ExplanationShow ActivationsShow Test FieldShow SteerShow Link

IFrame

Link

Negative Logits

üzel

-0.07

 amaç

-0.06

peating

-0.06

izens

-0.06

äge

-0.06

 совет

-0.06

 African

-0.06

звичай

-0.06

 racist

-0.06

 used

-0.06

POSITIVE LOGITS

 throne

0.16

 Throne

0.11

 Thrones

0.09

><?

0.07

(inputStream

0.07

0.07

◆

0.07

thro

0.07

坛

0.07

HONE

0.06

Activations Density 0.002%

No Known Activations