INDEX

Explanations

Improvement and critique

np_max-act · gemini-2.0-flash

discussions around software security practices and vulnerabilities.

oai_token-act-pair · gpt-4o-mini Triggered by @xinyanhu8

mentions of safety, security, risk, or warnings (words about danger, being safe, sanitize, threat/quality concerns).

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

language that evaluates quality, reliability, efficiency, correctness, or performance, often in the context of data integrity or requirements.

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

New Auto-Interp

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ازم

-0.06

.embedding

-0.06

 філ

-0.06

금

-0.06

Ist

-0.06

ILogger

-0.05

りに

-0.05

Between

-0.05

 بیرون

-0.05

シ

-0.05

POSITIVE LOGITS

................

0.08

	async

0.07

 organizing

0.07

getApplication

0.07

 Newsletter

0.07

 националь

0.07

 molecule

0.07

 molecules

0.07

........................

0.07

 alta

0.07

Activations Density 0.391%

Improvement and critique

discussions around software security practices and vulnerabilities.

mentions of safety, security, risk, or warnings (words about danger, being safe, sanitize, threat/quality concerns).

language that evaluates quality, reliability, efficiency, correctness, or performance, often in the context of data integrity or requirements.

No Comments

No Known Activations