INDEX

Explanations

specific years or numerical data associated with dates

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

yc

-0.18

elsing

-0.17

yll

-0.16

ilk

-0.16

alla

-0.15

zar

-0.15

ture

-0.15

shint

-0.15

aten

-0.14

unks

-0.14

POSITIVE LOGITS

ooter

0.14

-valu

0.14

istar

0.14

_BUSY

0.13

ichel

0.13

-context

0.13

 Ð´Ð¾ÑĢ

0.13

imuth

0.13

RIORITY

0.13

 tá»

0.13

Activations Density 0.001%

specific years or numerical data associated with dates

No Comments

No Known Activations