INDEX

Explanations

verbs related to giving or receiving something in return

oai_token-act-pair · gpt-3.5-turbo

terms related to reciprocity and corroboration

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

Khe

-0.97

GAME

-0.74

elf

-0.68

mire

-0.68

endor

-0.67

lace

-0.66

OWER

-0.62

dress

-0.61

lore

-0.61

mosp

-0.61

POSITIVE LOGITS

ating

1.14

ized

1.08

ated

1.06

ization

1.06

itive

1.04

izing

1.00

ity

0.91

ation

0.90

icum

0.84

ising

0.83

Activations Density 0.021%

verbs related to giving or receiving something in return

terms related to reciprocity and corroboration

No Comments

No Known Activations