INDEX

Explanations

words related to geographical locations or structures within a city

oai_token-act-pair · gpt-3.5-turbo

geographic and urban location references

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

pha

-0.74

ynthesis

-0.71

 damned

-0.68

orthy

-0.66

gat

-0.65

glomer

-0.64

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.64

 unlaw

-0.63

TAG

-0.61

ynt

-0.61

POSITIVE LOGITS

hesis

0.71

endment

0.67

alian

0.60

imester

0.60

 snack

0.59

Angelo

0.59

hyde

0.59

 showdown

0.59

çĶ°

0.59

Italian

0.58

Activations Density 0.146%

words related to geographical locations or structures within a city

geographic and urban location references

No Comments

No Known Activations