INDEX

Explanations

actions related to leaving or departing situations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 11-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L11R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.11.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

oid

-0.17

ameda

-0.15

ester

-0.15

ÐµÐ²Ð¸Ñĩ

-0.15

DOI

-0.15

ote

-0.14

ode

-0.14

agg

-0.14

cete

-0.14

atform

-0.14

POSITIVE LOGITS

-append

0.14

yz

0.14

izi

0.14

arakter

0.13

ylland

0.13

’n

0.13

äºķ

0.13

lodash

0.13

 Essen

0.13

bard

0.13

Activations Density 0.148%

actions related to leaving or departing situations

No Comments

No Known Activations