INDEX

Explanations

Assistant

np_max-act · gemini-2.0-flash

instructions and responses for a fact-checking QA task, especially assistant sections with “Reasoning” and a “Yes/No” verdict.

oai_token-act-pair · gpt-5 Triggered by @vetterc0

New Auto-Interp

Configuration

andyrdt/saes-qwen2.5-7b-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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ASS

-0.07

cart

-0.07

кат

-0.07

BarButton

-0.06

安东

-0.06

 Chips

-0.06

PY

-0.06

predicate

-0.06

肴

-0.06

交互

-0.06

POSITIVE LOGITS

 przez

0.08

 pelos

0.08

za

0.07

 Christina

0.07

Ryu

0.07

ulled

0.07

_advanced

0.07

 titanium

0.07

quez

0.07

日军

0.07

Activations Density 0.038%

Assistant

instructions and responses for a fact-checking QA task, especially assistant sections with “Reasoning” and a “Yes/No” verdict.

No Comments

No Known Activations