INDEX

Explanations

titles of TV shows and movies

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

dep

-0.15

kos

-0.14

Ä

-0.14

 guaranteed

-0.14

Cos

-0.14

 gang

-0.13

ederland

-0.13

 nerv

-0.13

 orth

-0.13

 garn

-0.13

POSITIVE LOGITS

_tE

0.18

_tA

0.17

_mE

0.17

nrw

0.15

_mD

0.15

_tF

0.15

_tD

0.15

ãģĭãģĹ

0.15

_mB

0.15

_tC

0.15

Activations Density 0.037%

titles of TV shows and movies

No Comments

No Known Activations