INDEX

Explanations

mentions of the term "News"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 27-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L27R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.27.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

ULE

-0.16

utton

-0.16

Å¼

-0.15

-0.14

tu

-0.14

inv

-0.14

tu

-0.14

Tu

-0.14

aring

-0.13

ÃŃt

-0.13

POSITIVE LOGITS

 Corp

0.20

day

0.19

week

0.18

aram

0.18

plex

0.17

busters

0.17

ela

0.16

ner

0.16

rud

0.16

ARAM

0.16

Activations Density 0.017%

mentions of the term "News"

No Comments

No Known Activations