INDEX

Explanations

proper nouns or names, particularly those related to people and possibly institutions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 6-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L6R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.6.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

udit

-0.15

 Lynch

-0.15

BarButton

-0.14

insky

-0.14

 æĿ

-0.14

ocker

-0.14

rido

-0.14

ãģĦãģ¤

-0.13

uffer

-0.13

 rein

-0.13

POSITIVE LOGITS

bish

0.16

kop

0.16

hire

0.15

 verd

0.15

bool

0.14

verity

0.14

yp

0.14

worth

0.13

dale

0.13

 sÃ©cur

0.13

Activations Density 0.539%

proper nouns or names, particularly those related to people and possibly institutions

No Comments

No Known Activations