INDEX

Explanations

category labels or classifications in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Ø§Ø°

-0.15

ulla

-0.15

gae

-0.15

ÑĮÐ²

-0.15

.vendor

-0.15

ocr

-0.15

oose

-0.14

 cá»©ng

-0.14

ening

-0.14

ragen

-0.14

POSITIVE LOGITS

HeaderValue

0.17

adox

0.16

ader

0.16

aks

0.15

CPF

0.15

bir

0.14

sez

0.14

 tang

0.14

è½

0.14

ud

0.14

Activations Density 0.005%

category labels or classifications in the text

No Comments

No Known Activations