INDEX

Explanations

words related to registration or being officially recorded

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

gesch

-0.16

ille

-0.15

sworth

-0.15

ington

-0.15

Ä±nÄ±

-0.14

.ide

-0.14

ikk

-0.14

igo

-0.14

ward

-0.14

grass

-0.14

POSITIVE LOGITS

/lic

0.22

_globals

0.19

/log

0.16

edly

0.16

horn

0.16

ISTRATION

0.16

/register

0.16

ificance

0.15

lands

0.15

ocene

0.15

Activations Density 0.026%

words related to registration or being officially recorded

No Comments

No Known Activations