INDEX

Explanations

the conjunction "and" indicating connections or additional thoughts in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

anes

-0.16

 à¤īà¤ł

-0.14

rica

-0.14

oust

-0.14

iano

-0.14

lexport

-0.14

CTIONS

-0.14

Ð³Ðµ

-0.14

Ð¾Ð²Ð°ÑĢ

-0.13

lobber

-0.13

POSITIVE LOGITS

 zwar

0.17

oulos

0.16

czy

0.15

ãģŁãģĦ

0.15

ATUS

0.15

rez

0.14

 ãĥĢ

0.14

iswa

0.13

ÏĦÏĥ

0.13

 TObject

0.13

Activations Density 0.055%

the conjunction "and" indicating connections or additional thoughts in the text

No Comments

No Known Activations