INDEX

Explanations

medical terms or linguistic leftovers that may be associated with medical conditions

oai_token-act-pair · gpt-3.5-turbo

entities related to formal groups, roles, and classifications

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

vous

-0.77

nox

-0.67

Sensor

-0.66

liest

-0.65

orthern

-0.64

arted

-0.63

consuming

-0.63

ftime

-0.62

 Galile

-0.61

aver

-0.61

POSITIVE LOGITS

 syndrome

0.74

(>

0.64

ryce

0.63

utsche

0.61

æ

0.61

å

0.58

odies

0.58

Qi

0.58

Tai

0.57

çĭ

0.57

Activations Density 0.728%

medical terms or linguistic leftovers that may be associated with medical conditions

entities related to formal groups, roles, and classifications

No Comments

No Known Activations