INDEX

Explanations

comparisons and similes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

iam

-0.16

_clr

-0.14

haar

-0.14

/Internal

-0.14

afort

-0.14

Ã¶h

-0.13

ochen

-0.13

leck

-0.13

ONTAL

-0.13

 somehow

-0.13

POSITIVE LOGITS

-than

0.17

than

0.16

 than

0.15

ately

0.15

Than

0.14

Ĥ¬

0.14

shint

0.14

abad

0.14

_wire

0.14

ddl

0.14

Activations Density 0.036%

comparisons and similes

No Comments

No Known Activations