INDEX

Explanations

repeating keywords and phrases throughout the document

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

.createFrom

-0.16

ÏĦÎ¹ÏĤ

-0.14

ãĥ³ãĥĪ

-0.14

.cgi

-0.14

 ÐĵÐ¾ÑģÐ¿

-0.14

 norm

-0.14

.Apis

-0.13

stry

-0.13

anie

-0.13

ionage

-0.13

POSITIVE LOGITS

^.

0.16

 Hook

0.15

constitutional

0.14

ossa

0.14

ÏĥÏī

0.14

egal

0.13

bjerg

0.13

è·

0.13

æ³ķ

0.13

 constitutional

0.13

Activations Density 0.009%

repeating keywords and phrases throughout the document

No Comments

No Known Activations