INDEX

Explanations

words related to specific brands or references in a context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

inem

-0.16

anca

-0.14

oid

-0.14

ilyn

-0.14

Assembler

-0.13

 Ð³ÑĢÐ¸

-0.13

olk

-0.13

íı°

-0.13

ansion

-0.13

Ø¶Ø§Ø¡

-0.13

POSITIVE LOGITS

 Ð¸Ñģ

0.16

eland

0.15

åľĪ

0.15

iga

0.14

.land

0.14

 Hutchinson

0.13

 durum

0.13

Î¾Î·

0.13

idas

0.13

usz

0.13

Activations Density 0.035%

words related to specific brands or references in a context

No Comments

No Known Activations