© 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 · 7-RESID-POST-AA · 30768 ｜ Neuronpedia

Home
Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
Llama3.1-8B-IT (Instruct)
Resid Post - 131k
7-RESID-POST-AA
30768

INDEX

Explanations

($

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_7/trainer_1

Dataset (Dashboard)

Various

Features

131,072

Data Type

float32

Hook Name

blocks.7.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

BBC

-0.07

'<

-0.07

 entity

-0.07

_rel

-0.07

.proj

-0.07

 Flem

-0.07

 vagy

-0.07

_stat

-0.06

REC

-0.06

 retains

-0.06

POSITIVE LOGITS

manın

0.07

ayın

0.07

($

0.07

$?

0.06

 хвор

0.06

 Woody

0.06

 Бог

0.06

.setTextColor

0.06

 reinforcements

0.06

 specializes

0.06

Activations Density 0.002%

No Known Activations