INDEX

Explanations

references to specific web links and identifiers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

akov

-0.16

ober

-0.15

par

-0.15

oser

-0.14

lig

-0.14

 adolescence

-0.14

oir

-0.14

æŀľ

-0.14

borrow

-0.14

ffi

-0.13

POSITIVE LOGITS

ÑģÑĤÐ°

0.17

inode

0.16

slaught

0.15

paque

0.15

ç¥¥

0.15

#ab

0.14

ushima

0.14

GMEM

0.14

roperties

0.14

ntax

0.14

Activations Density 0.030%

references to specific web links and identifiers

No Comments

No Known Activations