INDEX

Explanations

references to historical places or landmarks

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

/inet

-0.15

 ká»³

-0.14

ropy

-0.14

 Reserve

-0.14

.fn

-0.14

/^[

-0.14

ardon

-0.13

Ã¡Äį

-0.13

 Content

-0.13

Ã¡tek

-0.13

POSITIVE LOGITS

 settlement

0.20

 earliest

0.19

 Strateg

0.16

 later

0.16

hab

0.16

 attest

0.16

 settlements

0.16

 evid

0.15

 local

0.15

 feud

0.15

Activations Density 0.133%

references to historical places or landmarks

No Comments

No Known Activations