© 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 · 106637 ｜ Neuronpedia

Home
Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
Llama3.1-8B-IT (Instruct)
Resid Post - 131k
15-RESID-POST-AA
106637

INDEX

Explanations

ToString

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

_Action

-0.07

ROT

-0.06

(stop

-0.06

nip

-0.06

/tmp

-0.06

 hath

-0.06

anggal

-0.06

Construction

-0.06

 refusal

-0.06

jo

-0.06

POSITIVE LOGITS

 پي

0.07

 المللی

0.06

міні

0.06

incipal

0.06

vern

0.06

 DateFormatter

0.06

UpInside

0.06

.addWidget

0.06

습

0.06

ักเร

0.05

Activations Density 0.012%

No Known Activations