INDEX

Explanations

references to academic publications and journal articles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 25-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L25R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.25.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

DSA

-0.15

Ø¸Ùģ

-0.15

xo

-0.14

ause

-0.14

 Wooden

-0.14

 Bout

-0.14

uards

-0.14

Ð±Ð¾ÑĢÐ°

-0.14

 Aires

-0.13

POSITIVE LOGITS

ëħĦ

0.21

 ÙħÛĮÙĦØ§Ø¯ÛĮ

0.17

å¹´

0.17

cab

0.16

templ

0.16

ATED

0.16

æı

0.15

Bos

0.15

eko

0.14

Activations Density 0.045%

references to academic publications and journal articles

No Comments

No Known Activations