INDEX

Explanations

focus words related to support and promotion

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

nap

-0.14

fleet

-0.14

.ua

-0.14

Imp

-0.14

ÑĤÐ°Ð¶

-0.14

'gc

-0.14

 canonical

-0.14

Č↵

-0.14

 Canonical

-0.14

ÏĢÎ¿Ïį

-0.14

POSITIVE LOGITS

 above

0.17

 Above

0.16

above

0.15

 oben

0.15

prung

0.15

kom

0.14

Above

0.14

.AutoComplete

0.14

uppy

0.14

Ø§ÙģØ±

0.14

Activations Density 0.022%

focus words related to support and promotion

No Comments

No Known Activations