INDEX

Explanations

names related to "Ho"

oai_token-act-pair · gpt-3.5-turbo

references to the term "Ho" in various contexts

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

 confir

-0.85

 totality

-0.82

*/(

-0.81

CTR

-0.74

CPC

-0.72

 simul

-0.68

ä»£

-0.67

ãĤ¼ãĤ¦ãĤ¹

-0.65

 shortage

-0.65

 mileage

-0.64

POSITIVE LOGITS

oney

1.10

arding

1.08

ppy

1.06

LLOW

1.06

gey

1.02

pper

1.00

ppe

0.97

oves

0.97

pps

0.96

eful

0.96

Activations Density 0.010%

names related to "Ho"

references to the term "Ho" in various contexts

No Comments

No Known Activations