INDEX

Explanations

problem/issue

np_max-act · gemini-2.0-flash

The neuron fires on the introductory solution word in answer summaries (e.g. “solved” or “fixed”).

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

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

onec

-0.06

.OK

-0.06

iet

-0.06

 венти

-0.06

|%

-0.06

 Plato

-0.06

 Swiss

-0.06

_scalar

-0.06

:j

-0.06

Shr

-0.06

POSITIVE LOGITS

-muted

0.08

 unaware

0.07

(savedInstanceState

0.07

_begin

0.07

 gön

0.07

 Titles

0.06

ull

0.06

 donate

0.06

 distributed

0.06

比赛

0.06

Activations Density 0.015%

problem/issue

The neuron fires on the introductory solution word in answer summaries (e.g. “solved” or “fixed”).

No Comments

No Known Activations