INDEX

Explanations

The neuron activates for punctuation marks, especially periods at the end of paragraphs or sentences.

oai_token-act-pair · claude-3-7-sonnet-20250219 Triggered by @trainjumperx

assertions that indicate certainty or emphasis in a statement

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 5-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L5R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.5.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ersh

-0.15

agan

-0.14

uff

-0.14

wal

-0.14

 Barn

-0.14

que

-0.14

 Briggs

-0.13

eca

-0.13

avs

-0.13

 interrupted

-0.13

POSITIVE LOGITS

erb

0.16

.setViewport

0.16

cud

0.15

anken

0.15

essen

0.15

 Kraj

0.14

.elapsed

0.14

AccessType

0.14

 dispatch

0.14

 Grat

0.14

Activations Density 0.131%

The neuron activates for punctuation marks, especially periods at the end of paragraphs or sentences.

assertions that indicate certainty or emphasis in a statement

No Comments

No Known Activations