INDEX

Explanations

the word "is" and its variations, as well as forms of "bis" and "ism" indicating identity or state

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L1R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.1.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

èı

-0.15

initely

-0.15

 Huffman

-0.15

idy

-0.14

olin

-0.14

}->{

-0.14

plementation

-0.13

dale

-0.13

 Transitional

-0.13

POSITIVE LOGITS

alue

0.17

istÃ©

0.15

Accessor

0.15

/tos

0.15

ilos

0.15

berger

0.15

çĴ°

0.14

ÄĻd

0.14

ÙĦÙĥ

0.14

ivot

0.14

Activations Density 0.020%

the word "is" and its variations, as well as forms of "bis" and "ism" indicating identity or state

No Comments

No Known Activations