INDEX

Explanations

negative qualifiers related to expectations and durations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 7-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L7R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.7.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

EEDED

-0.16

indle

-0.16

uco

-0.15

asic

-0.14

anga

-0.14

Î±ÏģÎ±

-0.14

ipple

-0.14

Affected

-0.14

aras

-0.14

POSITIVE LOGITS

BackColor

0.14

 temporal

0.14

 ç»ı

0.14

ritch

0.14

ystore

0.14

Temporal

0.14

ÏĥÏĥ

0.13

Sud

0.13

parms

0.13

ê¼

0.13

Activations Density 0.333%

negative qualifiers related to expectations and durations

No Comments

No Known Activations