INDEX

Explanations

historical or geographical references to cities and fortifications

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

quez

-0.18

ajes

-0.18

@student

-0.16

ridged

-0.15

kova

-0.15

reative

-0.15

è¦ª

-0.15

byss

-0.15

ungan

-0.14

ilan

-0.14

POSITIVE LOGITS

Ĥ¬

0.17

onio

0.15

(BitConverter

0.14

Ã©ro

0.14

defgroup

0.14

Ryu

0.14

æ¯

0.14

 involved

0.14

ħį

0.13

 Ð²ÐºÑĥ

0.13

Activations Density 0.075%

historical or geographical references to cities and fortifications

No Comments

No Known Activations