INDEX

Explanations

think

np_max-act · gemini-2.0-flash

The neuron fires on tokens that signal an authorial opinion or hedging—e.g. “think” (as in “We think”)—marking subjective or speculative language.

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

urum

-0.07

 outfit

-0.07

 mask

-0.06

 \\↵

-0.06

 DecimalFormat

-0.06

 radical

-0.06

Src

-0.06

.SOCK

-0.06

 preset

-0.06

POSITIVE LOGITS

 handleMessage

0.07

Traditional

0.06

igrate

0.06

Oy

0.06

Tanggal

0.06

.randn

0.06

ướ

0.06

_TIM

0.06

vají

0.06

혀

0.06

Activations Density 0.001%

think

The neuron fires on tokens that signal an authorial opinion or hedging—e.g. “think” (as in “We think”)—marking subjective or speculative language.

No Comments

No Known Activations