INDEX

Explanations

specific non-English characters and symbols, indicating a focus on content in a different language or encoding

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 13-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L13R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.13.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

TK

-0.16

endor

-0.16

 Vere

-0.15

ing

-0.15

 Couch

-0.14

Eh

-0.14

 Curl

-0.14

eye

-0.14

aved

-0.14

avanaugh

-0.14

POSITIVE LOGITS

addir

0.18

onaut

0.17

ÐĲÑĢÑħÑĸÐ²

0.16

 realized

0.15

.gstatic

0.15

oje

0.14

@brief

0.14

Drag

0.14

rÃ³d

0.14

ndata

0.13

Activations Density 0.070%

specific non-English characters and symbols, indicating a focus on content in a different language or encoding

No Comments

No Known Activations