INDEX

Explanations

words related to filtering and filters

oai_token-act-pair · gpt-3.5-turbo

words related to filtering processes or materials

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

eanor

-0.84

ciating

-0.74

olars

-0.72

ensional

-0.70

nee

-0.69

roach

-0.68

lished

-0.68

peak

-0.68

rious

-0.68

ILY

-0.67

POSITIVE LOGITS

 filters

0.93

 filter

0.85

filter

0.84

 filtering

0.84

 Filter

0.84

 cutoff

0.82

 sweeps

0.76

operator

0.75

Enlarge

0.71

 curtains

0.69

Activations Density 0.015%

words related to filtering and filters

words related to filtering processes or materials

No Comments

No Known Activations