INDEX

Explanations

structure and syntax within code

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ervas

-0.17

iday

-0.15

Ala

-0.15

yo

-0.15

vit

-0.14

erras

-0.14

à¤¿à¤¯à¤¤

-0.14

inois

-0.13

utions

-0.13

CLAIM

-0.13

POSITIVE LOGITS

 foreach

0.17

	foreach

0.16

loquent

0.16

foreach

0.15

SWG

0.15

()->

0.15

rox

0.15

Hab

0.15

 Bowman

0.15

æĹıèĩªæ²»

0.15

Activations Density 0.020%

structure and syntax within code

No Comments

No Known Activations