© 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 · 52163 ｜ Neuronpedia

Home
Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
Llama3.1-8B-IT (Instruct)
Resid Post - 131k
11-RESID-POST-AA
52163

INDEX

Explanations

arrow symbols

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_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

correct

-0.08

(service

-0.07

anic

-0.07

gol

-0.06

_trees

-0.06

lexer

-0.06

 abound

-0.06

_CLK

-0.06

 drilled

-0.06

Misc

-0.06

POSITIVE LOGITS

 outputPath

0.07

 رج

0.07

0.06

_SLOT

0.06

wind

0.06

 이동

0.06

,SLOT

0.06

 щодо

0.06

ByVersion

0.06

ровать

0.06

Activations Density 0.003%

No Known Activations