INDEX

Explanations

comparisons or similes

oai_token-act-pair · gpt-3.5-turbo

variations of the word "like" and related forms indicating comparison or resemblance

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

Nou

-0.76

ACA

-0.76

 systematic

-0.68

 Schools

-0.67

 residency

-0.66

INA

-0.66

Interstitial

-0.65

 Techniques

-0.65

trl

-0.63

 fragment

-0.62

POSITIVE LOGITS

lihood

1.57

lik

1.10

liness

0.93

lik

0.91

ened

0.90

iour

0.88

ability

0.87

ening

0.86

 likeness

0.82

liest

0.80

Activations Density 0.004%

comparisons or similes

variations of the word "like" and related forms indicating comparison or resemblance

No Comments

No Known Activations