INDEX

Explanations

URLs or links to web content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

FIG

-0.16

orex

-0.14

 impulse

-0.14

Fig

-0.14

cano

-0.14

 collar

-0.14

-0.13

 tenure

-0.13

_patterns

-0.13

POSITIVE LOGITS

.twitter

0.19

_via

0.17

pic

0.17

amax

0.17

Via

0.17

icare

0.17

via

0.16

Via

0.16

via

0.15

OMATIC

0.15

Activations Density 0.006%

URLs or links to web content

No Comments

No Known Activations