INDEX

Explanations

code-related syntax and structure elements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 22-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L22R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.22.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

 underst

-0.16

unate

-0.14

 Carm

-0.14

odox

-0.14

UPS

-0.14

 Terrain

-0.14

tring

-0.14

 sant

-0.13

modern

-0.13

POSITIVE LOGITS

lej

0.15

ãĥĥãĥĦ

0.15

asher

0.15

ilogy

0.14

ij

0.14

IFA

0.14

acro

0.14

Pey

0.14

ÏĨÎ·

0.13

UNCH

0.13

Activations Density 0.005%

code-related syntax and structure elements

No Comments

No Known Activations