INDEX

Explanations

discussions around academic scholarship and critiquing historical interpretations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 20-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L20R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.20.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

OK

-0.15

 Philosoph

-0.15

 assort

-0.14

 Roles

-0.14

 impress

-0.14

 Philosophy

-0.14

 philosoph

-0.14

ÏģÏĮÏĤ

-0.14

itemid

-0.14

iesz

-0.13

POSITIVE LOGITS

 CrossAxisAlignment

0.17

çłĶç©¶

0.15

 neob

0.15

iban

0.14

nio

0.14

usan

0.14

nist

0.14

comings

0.14

voices

0.14

Activations Density 0.091%

discussions around academic scholarship and critiquing historical interpretations

No Comments

No Known Activations