INDEX

Explanations

phrases related to providing reasons or justifications

oai_token-act-pair · gpt-3.5-turbo

phrases that reference criteria or reasons for judgments and decisions

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

ascus

-0.73

aspers

-0.67

rys

-0.66

stall

-0.65

knit

-0.65

ortium

-0.64

nova

-0.64

byter

-0.63

nery

-0.60

jer

-0.60

POSITIVE LOGITS

 assumption

0.83

 occasions

0.75

 either

0.72

 premise

0.69

 basis

0.66

 behalf

0.65

 footing

0.65

 grounds

0.64

 whim

0.64

GUI

0.64

Activations Density 0.039%

phrases related to providing reasons or justifications

phrases that reference criteria or reasons for judgments and decisions

No Comments

No Known Activations