INDEX

Explanations

references to notions of "the best" or "greatest" in a context of comparison or ranking

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.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

ÐµÐ»Ð¸Ñĩ

-0.18

ynamic

-0.15

Ø§ÙĬØ±

-0.15

ãĥĥãĥĹ

-0.14

ëĦ

-0.14

ibold

-0.14

missive

-0.14

elay

-0.13

shal

-0.13

ifferential

-0.13

POSITIVE LOGITS

 ages

0.18

 time

0.17

brit

0.15

ignment

0.15

wed

0.15

otts

0.15

 Ages

0.15

nehmer

0.15

ridge

0.14

igator

0.14

Activations Density 0.011%

references to notions of "the best" or "greatest" in a context of comparison or ranking

No Comments

No Known Activations