INDEX

Explanations

.

np_max-act · gemini-2.0-flash

The neuron detects the mention of “tools,” especially when the text is instructing or listing software/utilities.

oai_token-act-pair · o4-mini Triggered by @xinyanhu8

sections of technical or structured content — e.g., headings, numbered lists, code/SQL blocks and form-like metadata.

oai_token-act-pair · gpt-5-mini 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

ері

-0.06

amate

-0.06

/H

-0.06

vt

-0.06

 pasar

-0.06

qh

-0.06

 Cần

-0.06

 아니라

-0.06

theless

-0.06

POSITIVE LOGITS

 parç

0.07

 де

0.07

odi

0.07

lied

0.07

파

0.07

 ratified

0.06

rpc

0.06

_OR

0.06

ΟΥ

0.06

 getModel

0.06

Activations Density 0.351%

.

The neuron detects the mention of “tools,” especially when the text is instructing or listing software/utilities.

sections of technical or structured content — e.g., headings, numbered lists, code/SQL blocks and form-like metadata.

No Comments

No Known Activations

.

The neuron detects the mention of “tools,” especially when the text is instructing or listing software/utilities.

sections of technical or structured content — e.g., headings, numbered lists, code/SQL blocks and form-like metadata.

No Comments

No Known Activations