INDEX

Explanations

references to code structures and programming concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

 adjustable

-0.15

 Millenn

-0.15

stadt

-0.15

uite

-0.15

 FIXED

-0.15

ÃŃstica

-0.14

 McCart

-0.14

zap

-0.14

urette

-0.14

.bootstrap

-0.14

POSITIVE LOGITS

_simps

0.16

 Gravity

0.16

Pun

0.15

odega

0.15

 reim

0.14

 Orange

0.14

 Contributor

0.14

intern

0.14

 gravity

0.14

affles

0.14

Activations Density 0.030%

references to code structures and programming concepts

No Comments

No Known Activations