© Neuronpedia 2026

Privacy & Terms Blog GitHub Slack Twitter Contact

Neuronpedia

Natural Language

NEW Assistant AxisNEW Circuit TracerUPDATESteer SAE Evals ExportsAPI Community Blog Privacy & Terms Contact

QWEN2.5-7B-IT · 23-RESID-POST-AA · 1191 ｜ Neuronpedia

Home
Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
Qwen2.5-7B-IT
Resid Post - 131k
23-RESID-POST-AA
1191

INDEX

Explanations

multidimensional concepts

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Top Features by Cosine Similarity

Configuration

andyrdt/saes-qwen2.5-7b-instruct/resid_post_layer_23/trainer_1

Dataset (Dashboard)

Various

Features

131,072

Data Type

float32

Hook Name

blocks.23.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

恫

-0.08

_permission

-0.07

صاحب

-0.07

ۥ

-0.07

 Dealer

-0.07

.PackageManager

-0.07

襞

-0.07

廓

-0.07

佝

-0.07

✩

-0.07

POSITIVE LOGITS

(USER

0.07

opian

0.07

lyph

0.07

 murders

0.07

+");↵

0.06

NUMBER

0.06

设定

0.06

 observing

0.06

纪念

0.06

负面

0.06

Activations Density 0.125%

No Known Activations