INDEX

Explanations

commitments to action or promises of improvement

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 5-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L5R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.5.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

smith

-0.14

akah

-0.14

IMA

-0.14

qd

-0.14

ãĥ¬ãĥ³

-0.14

 subst

-0.14

yne

-0.14

agged

-0.13

HttpException

-0.13

 poly

-0.13

POSITIVE LOGITS

icana

0.16

_AN

0.15

 exhaust

0.15

rzy

0.15

 ÑģÐ¿Ð¾ÑĢ

0.14

NZ

0.14

 Clips

0.14

erez

0.14

(&_

0.14

ARAM

0.14

Activations Density 0.144%

commitments to action or promises of improvement

No Comments

No Known Activations