INDEX

Explanations

numerical sequences or identifiers often associated with data or timestamps

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

hlas

-0.16

elles

-0.15

onas

-0.14

 Bair

-0.14

egade

-0.14

sworth

-0.14

illow

-0.14

ÑĨÐ°ÑĤÑĮ

-0.13

covered

-0.13

 OPTIONAL

-0.13

POSITIVE LOGITS

Å¯r

0.15

]|[

0.15

Ð»ÑĥÐ±

0.14

 prolet

0.14

abi

0.14

 Quarry

0.13

argins

0.13

obuf

0.13

 Wheels

0.13

Activations Density 0.030%

numerical sequences or identifiers often associated with data or timestamps

No Comments

No Known Activations