INDEX

Explanations

arguments

np_max-act · gemini-2.0-flash

The neuron lights up on the token sequence for "complex" / "complexity" (and related morphology), i.e., words signaling complexity.

oai_token-act-pair · gpt-5-mini 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

彝

-0.08

 prefetch

-0.08

-0.07

)`

-0.07

召唤

-0.07

ს

-0.07

泥

-0.07

 eBooks

-0.06

_serv

-0.06

ULD

-0.06

POSITIVE LOGITS

已有

0.07

ayaran

0.07

idable

0.07

 Remaining

0.07

明顯

0.07

Throwable

0.07

 fácil

0.07

閱

0.07

/>↵

0.06

的质量

0.06

Activations Density 0.285%

arguments

The neuron lights up on the token sequence for "complex" / "complexity" (and related morphology), i.e., words signaling complexity.

No Comments

No Known Activations

arguments

The neuron lights up on the token sequence for "complex" / "complexity" (and related morphology), i.e., words signaling complexity.

No Comments

No Known Activations