INDEX

Explanations

experiences or actions related to personal history or encounters

oai_token-act-pair · gpt-3.5-turbo

experiences and actions related to personal skills and encounters

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

dan

-0.67

cms

-0.67

ju

-0.64

BP

-0.61

 Supplemental

-0.60

Accessory

-0.56

 upstream

-0.55

Gameplay

-0.55

np

-0.54

 sinks

-0.54

POSITIVE LOGITS

 anywhere

0.81

 EVER

0.78

nor

0.77

 dime

0.73

ukong

0.72

 slightest

0.68

 adolesc

0.67

 remotely

0.64

ãĤ¼ãĤ¦ãĤ¹

0.63

 Pastebin

0.62

Activations Density 0.371%

experiences or actions related to personal history or encounters

experiences and actions related to personal skills and encounters

No Comments

No Known Activations