INDEX

Explanations

references to architectural features and structural descriptions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 27-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L27R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.27.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

uess

-0.15

ÅĤaw

-0.15

 Marsh

-0.14

 Dunn

-0.14

 cast

-0.14

/tty

-0.14

OLOR

-0.14

orsch

-0.14

ICAST

-0.14

rong

-0.14

POSITIVE LOGITS

 later

0.16

later

0.15

 Later

0.15

oplast

0.15

ÑģÐ¾

0.15

 tapi

0.15

disposed

0.14

 disposition

0.14

 constructive

0.14

 dispos

0.14

Activations Density 0.028%

references to architectural features and structural descriptions

No Comments

No Known Activations