INDEX

Explanations

conjunctions that often appear in a list format

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

_Tis

-0.16

OffsetTable

-0.15

nes

-0.14

 Boyle

-0.14

\Collections

-0.14

åĳ¨æľŁ

-0.14

ÙĪØ¬Ùĩ

-0.14

 Ð´ÑĢ

-0.14

ãĢ

-0.14

podob

-0.13

POSITIVE LOGITS

/or

0.19

and

0.14

 bedding

0.14

loi

0.14

 doubt

0.14

but

0.14

ijd

0.14

 directly

0.14

AGON

0.13

fect

0.13

Activations Density 0.181%

conjunctions that often appear in a list format

No Comments

No Known Activations

conjunctions that often appear in a list format

No Comments

No Known Activations