INDEX

Explanations

phrases that indicate exceptions or notable distinctions in situations

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

eno

-0.18

ral

-0.17

thon

-0.16

iol

-0.16

orte

-0.16

annah

-0.15

end

-0.15

isa

-0.15

iom

-0.15

no

-0.15

POSITIVE LOGITS

 hÃ£

0.15

DTV

0.15

stad

0.14

æĬĺ

0.14

Equals

0.14

 exceptions

0.14

istar

0.14

)did

0.14

modifiable

0.14

_tokenize

0.14

Activations Density 0.282%

phrases that indicate exceptions or notable distinctions in situations

No Comments

No Known Activations