INDEX

Explanations

references to sports events, teams, and scores

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 11-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L11R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.11.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

conomy

-0.17

abin

-0.16

sled

-0.15

je

-0.14

 láº½

-0.13

rus

-0.13

ERSIST

-0.13

bes

-0.13

ebin

-0.13

afs

-0.13

POSITIVE LOGITS

 Pied

0.15

outers

0.15

 Rank

0.14

ritel

0.14

casting

0.14

ORK

0.14

uc

0.14

WhiteSpace

0.13

Activations Density 0.005%

references to sports events, teams, and scores

No Comments

No Known Activations