INDEX

Explanations

titles, names, or identifiers related to various creative works or industry terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 7-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L7R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.7.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

gree

-0.15

iá»ĩt

-0.14

}elseif

-0.14

Ø±Ø³ÛĮ

-0.14

verse

-0.14

ÏģÏĮ

-0.14

uner

-0.14

ekim

-0.14

cone

-0.14

oucher

-0.14

POSITIVE LOGITS

Ð°ÑĢÑĸÐ²

0.16

inde

0.15

_exempt

0.14

 ÑĢÑı

0.14

statt

0.14

soever

0.14

ORAGE

0.14

Ð°Ð½Ð¸ÑĨ

0.14

 phÃ©p

0.13

Activations Density 0.971%

titles, names, or identifiers related to various creative works or industry terms

No Comments

No Known Activations