© 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 · 75475 ｜ Neuronpedia

Home
Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
Llama3.1-8B-IT (Instruct)
Resid Post - 131k
15-RESID-POST-AA
75475

INDEX

Explanations

Poland

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

 drilled

-0.06

forgot

-0.06

бург

-0.06

ンチ

-0.06

 urges

-0.06

-0.06

 Пос

-0.06

 nécessaire

-0.06

ВА

-0.06

Det

-0.06

POSITIVE LOGITS

_SYMBOL

0.07

_proba

0.07

 robot

0.06

 sélection

0.06

_Copy

0.06

\P

0.06

.assertNot

0.06

-answer

0.06

 тот

0.06

.serialization

0.06

Activations Density 0.008%

No Known Activations