INDEX

Explanations

words and phrases that indicate conditional scenarios or correlations in reasoning and argumentation

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

Validator

-0.14

hir

-0.14

 ÃĩaÄŁ

-0.14

.returnValue

-0.14

.CONFIG

-0.14

agle

-0.14

-0.13

à¤°à¤¤

-0.13

 payer

-0.13

 Claus

-0.13

POSITIVE LOGITS

bdb

0.15

andal

0.15

itra

0.14

à¤¾à¤¯à¤¨

0.14

iman

0.14

iais

0.14

 enrichment

0.14

onnen

0.14

ifar

0.14

rito

0.13

Activations Density 0.001%

words and phrases that indicate conditional scenarios or correlations in reasoning and argumentation

No Comments

No Known Activations