INDEX

Explanations

programming-related constructs, particularly those related to file names and data structures

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

EGA

-0.17

ÏĢÎ±

-0.16

XS

-0.15

geist

-0.15

à¤²à¤¤

-0.15

.Constraint

-0.15

ABCDEFGHI

-0.15

RuleContext

-0.15

smr

-0.15

íıŃ

-0.14

POSITIVE LOGITS

 Deng

0.15

atan

0.14

 McKenzie

0.14

IS

0.14

 Responsive

0.14

ore

0.14

orn

0.13

ÑĤÑĮ

0.13

co

0.13

Activations Density 0.009%

programming-related constructs, particularly those related to file names and data structures

No Comments

No Known Activations