INDEX

Explanations

programming symbols and structure in code snippets

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

@nate

-0.17

 Meth

-0.16

iron

-0.16

SOLE

-0.15

olik

-0.15

ÐĲÑĢÑħÑĸÐ²

-0.15

fir

-0.14

IRON

-0.14

ÑĥÑĩ

-0.14

olson

-0.14

POSITIVE LOGITS

ahan

0.17

axon

0.16

rete

0.15

ÅĻÃŃ

0.14

ror

0.14

Yan

0.14

aves

0.14

rve

0.13

 stepper

0.13

rons

0.13

Activations Density 0.814%

programming symbols and structure in code snippets

No Comments

No Known Activations