INDEX

Explanations

references to programming namespaces and classes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

ÙĪØ§ÙĨ

-0.15

amd

-0.15

 sigmoid

-0.14

è½®

-0.14

_mE

-0.14

oust

-0.14

èī

-0.14

ä½³

-0.14

rt

-0.13

anza

-0.13

POSITIVE LOGITS

ONENT

0.16

 bore

0.16

emet

0.15

ButtonDown

0.15

 Pall

0.15

emade

0.15

æľ

0.14

 âĹĦ

0.14

izzo

0.14

##_

0.14

Activations Density 0.001%

references to programming namespaces and classes

No Comments

No Known Activations