INDEX

Explanations

words related to comments and commenting behavior

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

ucha

-0.15

pel

-0.15

inel

-0.15

ouz

-0.14

chet

-0.14

andest

-0.14

abit

-0.14

yo

-0.14

 Silent

-0.14

coma

-0.14

POSITIVE LOGITS

aries

0.19

ìĤ¬íķŃ

0.16

/Instruction

0.16

eting

0.16

ICTURE

0.16

lint

0.15

ariat

0.15

 ìĤ¬íķŃ

0.15

ary

0.15

ers

0.14

Activations Density 0.033%

words related to comments and commenting behavior

No Comments

No Known Activations