INDEX

Explanations

phrases related to asking for or giving explanations or instructions

oai_token-act-pair · gpt-3.5-turbo

expressions indicating uncertainty or questions about value or worth

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

spir

-0.65

isp

-0.64

native

-0.62

ccoli

-0.62

 glyph

-0.59

£ı

-0.59

 undert

-0.59

 migr

-0.58

 Janeiro

-0.57

 footh

-0.57

POSITIVE LOGITS

Answer

0.98

 Nope

0.92

?????-?????-

0.87

 Answer

0.83

.?

0.82

 answered

0.71

Reference

0.68

?,

0.67

 voucher

0.67

Ü

0.67

Activations Density 0.183%

phrases related to asking for or giving explanations or instructions

expressions indicating uncertainty or questions about value or worth

No Comments

No Known Activations