© Neuronpedia 2026

Privacy & Terms Blog GitHub Slack Twitter Contact

Neuronpedia

APIAssistant AxisNEW Circuit TracerNEW Steer SAE Evals Exports Community Blog Privacy & Terms Contact

LLAMA3.1-8B-IT · 15-RESID-POST-AA · 90796 ｜ Neuronpedia

Home
Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
Llama3.1-8B-IT (Instruct)
Resid Post - 131k
15-RESID-POST-AA
90796

INDEX

Explanations

abuse and bullying

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_15/trainer_1

Dataset (Dashboard)

Various

Features

131,072

Data Type

float32

Hook Name

blocks.15.hook_resid_post

Architecture

standard

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

Show PlotsShow ExplanationShow ActivationsShow Test FieldShow SteerShow Link

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Samoa

-0.07

 التص

-0.07

(My

-0.07

_Float

-0.07

(".")

-0.07

Her

-0.06

 Express

-0.06

Obama

-0.06

.worker

-0.06

.Cross

-0.06

POSITIVE LOGITS

 competed

0.07

dé

0.07

 stir

0.07

termin

0.07

sko

0.06

 Preparation

0.06

solid

0.06

binding

0.06

 connexion

0.06

 trú

0.06

Activations Density 0.022%

No Known Activations