INDEX

Explanations

terms related to quantifiable evaluations and comparisons

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

 Hick

-0.17

foon

-0.16

ichen

-0.15

ymb

-0.15

Tw

-0.14

ngo

-0.14

del

-0.14

obo

-0.14

è¶Ĭ

-0.14

POSITIVE LOGITS

oller

0.15

arro

0.15

ensi

0.15

 stake

0.14

sha

0.14

pling

0.14

 Stra

0.14

+-

0.14

undred

0.14

batis

0.14

Activations Density 0.002%

terms related to quantifiable evaluations and comparisons

No Comments

No Known Activations