INDEX

Explanations

the word "Ming" and its variations, potentially related to names or cultural references

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 2-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L2R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.2.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

WEBPACK

-0.17

-0.16

son

-0.16

 GOODS

-0.15

upon

-0.14

onso

-0.14

bard

-0.14

ÄĽÅ¾

-0.14

irates

-0.14

entin

-0.14

POSITIVE LOGITS

elm

0.16

 softer

0.16

osta

0.15

 coloured

0.15

ÏĥÎ¿

0.15

nock

0.15

enser

0.14

amu

0.14

ens

0.14

Ã¸r

0.14

Activations Density 0.026%

the word "Ming" and its variations, potentially related to names or cultural references

No Known Activations