INDEX

Explanations

various forms of the word "to" and related conditional or quantitative terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.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

rench

-0.15

affected

-0.14

ens

-0.14

ufen

-0.13

 wors

-0.13

ropolis

-0.13

Ð°Ð»ÑĸÐ·

-0.13

annis

-0.13

erp

-0.13

POSITIVE LOGITS

lut

0.14

referrer

0.14

reso

0.14

(æľĪ

0.14

oni

0.14

openh

0.13

-urlencoded

0.13

ÙĤ

0.13

trinsic

0.13

dra

0.13

Activations Density 0.012%

various forms of the word "to" and related conditional or quantitative terms

No Comments

No Known Activations