© 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 · 11-RESID-POST-AA · 13385 ｜ Neuronpedia

Home
Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
Llama3.1-8B-IT (Instruct)
Resid Post - 131k
11-RESID-POST-AA
13385

INDEX

Explanations

HTML code

np_max-act · gemini-2.0-flash

This neuron activates on HTML “div” tags.

oai_token-act-pair · o4-mini Triggered by @xinyanhu8

New Auto-Interp

Top Features by Cosine Similarity

Configuration

andyrdt/saes-llama-3.1-8b-instruct/resid_post_layer_11/trainer_1

Dataset (Dashboard)

Various

Features

131,072

Data Type

float32

Hook Name

blocks.11.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

 bern

-0.06

］

-0.06

>>

-0.06

аксим

-0.06

PAD

-0.06

ách

-0.06

숨

-0.06

 Deal

-0.06

ございます

-0.06

-0.05

POSITIVE LOGITS

表现

0.07

 conven

0.07

иж

0.07

ทำ

0.07

0.07

 đổi

0.07

blind

0.06

SCRIBE

0.06

\data

0.06

CURRENT

0.06

Activations Density 0.003%

No Known Activations