INDEX

Explanations

terms related to the action of processing

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

uchi

-0.16

indow

-0.15

rench

-0.14

crest

-0.14

ê¼

-0.14

Å¼

-0.14

tas

-0.14

yre

-0.14

kÃ¼

-0.14

POSITIVE LOGITS

/display

0.19

yonel

0.16

edImage

0.16

ible

0.16

edly

0.15

sing

0.15

uality

0.15

cla

0.15

/output

0.14

enko

0.14

Activations Density 0.030%

terms related to the action of processing

No Comments

No Known Activations